<<

This document is out of print and the writing is beyond copyright, or it is being reproduced with the permission of the authors and publishers. This version reflects the general appearance of the original, though modern computer-based type has been substituted with the consequence that within-page spacing is slightly altered. Pagination has been respected so that citations will remain correct. This electronic version of the work is copyrighted. You may reproduce and distribute it, but only for free and only with this notice intact. Related material is available at our web site.

©Fred Collopy 2004.

THE VISUAL NATURE OF

Into the same rivers we step and we do not step; we are and we are not.

Heraclitus

THE VISUAL NATURE OF COLOR

PATRICIA SLOANE

DESIGN PRESS

ACKNOWLEDGMENTS Research for this book was partially supported by a fellowship from the John Simon Guggenktelm Memorial Foundation, and the New York City Technical College of The City University of New York. For their suggestions and encouragement, I would like to thank Robert Motherwell, Rudolph Arnhelm, David Ecker, Jerome Hausman, Robert Ginsberg, Pamela Dohner, Linda Venator, Kurt Wildermuth, and Nancy .

First Edition, First Printing Copyright © 1989 by Patricia Sloane Printed in the United States of America Designed by Gilda Hannah

Reproduction or publication of the content in any manner, without express permission of the publisher, is prohibited. The publisher takes no responsibility for the use of any of the materials or methods described in this book, or for the products thereof.

Library of Congress Cataloging in Publication Data Sloane, Patricia. The visual nature of color. Bibliography: p. Includes Index. 1. Color in art. 2. Color (Philosophy) 3. Color- Psychological aspects. I. Title. ND1488.555 1989 701'.8 88-33624 ISBN 0-8036-5500-X

Design Press offers posters and The Cropper, a device for cropping artwork, for sale. For information, contact Mailorder Department. Design Press books are available at special discounts for bulk purchases for sales promotions, fund raisers, or premiums. For details contact Special Sales Manager. Questions regarding the content of the book should be addressed to: DESIGN PRESS Division of TAB BOOKS Inc. 10 East 21st Street New York, NY 10010

Contents

Preface vii

PART ONE: COLOR AND LANGUAGE 1. Learning to Use Color Names 3 2. Color as a Continuum 11 3. Understanding Color Names 17 4. The Limits of Language and the Logic of Color 26 5. Knowing How to Identify Color 3 7

PART TWO: COLOR AND 6. Light of Day, Dark of Night 47 7. Light and Dark in Perspective 55 8. Newton 64 9. The Cause of Color and Light 73 10. -, , , and Optical Mixture 79 11. Achromatic and Mirrors 88 12. Color Causes Light 97 13. Light as Symbol and Visual Metaphor I 106 14. Light as Symbol and Visual Metaphor II 114 15. Nonvisual Seeing and the Metameric Grays 126

PART THREE: COLOR AND FORM 16. The Two-Dimensional World 135 17. Delusion and the Geometry of Visual Space 143 18. Extension in the Visual Field 157 19. Complernentarity in the Visual Field 167 20. Color Fields and Colored Forms 175

PART FOUR: COLOR AND 21. , Color, and Culture 187 22. Prime Minister Gladstone and the 193 23. Bassa, Shona, and Ibo 201 24. Tristimulus Theory and 213 25. Color and Form in Art 224 26. Subjectivity and the Number of Colors 230 27. Object or Attribute 237 28. Conveying Information about Color 251 29. On Ambiguity in Color Names 262

PART FIVE: 30. Systematizers and Systems 2 71 31. The Logarithmic Gray Scale 281 32. Primary, Secondary, and Tertiary Colors 293 33. 301 34. Color Mixture 313 35. Additive/Subtractive Theory 323

Notes 329 Bibliography 332

Preface

Color and vision are inseparable. What we see We learn from the biological sciences that is composed of colors, though the question the colors of living organisms are functional, "what is color?" yields inconclusive answers. that life on earth would be different if the The study of color is not an academic dis- chlorophyll of plants was not green, if human cipline in its own right, and many disciplines blood was not red. Astronomy and cosmology, claim pieces of it. Physics owns the question of growing ever closer to particle physics, consider how color is caused, leaving to philosophy and what the colors of the stars and sky suggest psychology inquiry about whether color is about the origin of the universe and about the chimerical and how we interpret what we see. origin of colors. The relationship between color and form is The social sciences offer theories about color touched on in the philosophy of design. names, about the relationship between the Geometry tells us what forms are and hints at experiences of seeing color and the words used what color is not. to describe what we see. Clues to the early The issue of beauty or harmoniousness in history of human ideas about color are also color combinations, and whether the terms revealed by the etymology of key color terms, can be defined, are addressed in the writings the history of what these terms have meant and of artists and the literature of art education. their cognates-words from which they may have These questions also involve disciplines as been derived. diverse as aesthetics and . Finding the means to combine the theories

of these diverse disciplines is less the task of fragments, can the fragments be fit together to Sisyphus than that of Isis, collecting the make ? One conception of truth relies on fragments of the corpse of Osiris so that they the metaphor of the montage, derived from could be reassembled and brought back to life. cinematography. By this reasoning, each Who could be qualified for such a task? discipline regards color from its own point of Nobodyor anybody, because we all know what view, beginning from its own premises. Truth is we see. approached by imagining the several points of Those who have contributed to our view laid atop one another like a montage; the understanding of color or wrote at length about inquirer looks into the heart of the matter it represent these different disciplines. Robert through the layers of the montage. Thus, color is Boyle and Sir Isaac Newton regarded in one sense what the physicist tells us. In other themselves as natural philosophers, though we , it is what artists, philosophers, call them physicists. Johann Wolfgang von psychologists, and workers in other areas say. In Goethe was a poet. Thomas Young was a total, color is to be understood as an aggregate physician. M.E. Chevreul was a chemist who of all the points of view about it. served for a time as the director of a dye I find this manner of reasoning more factory. William Ewart Gladstone was Queen confusing than helpful. Color is a singular Victoria's prime minister. Albert H. Munsell phenomenon. If it is to be chopped into pieces was an art educator. Wilhelm Ostwald, who called points of view, the points of view must be shared Munsell's hope of simplifying the use of consistent within themselves and consistent with color for industry, began a second career as a one another. The pieces of the jigsaw puzzle color theorist after retiring as a professor of must fit together. We cannot assume that all physical chemistry. points of view are equal or that they all meet the Those interested in color should read tests of logical consistency and consistency with Vincent van Gogh's letters, Eugene Delacroix's what we see. journals, and the many statements and Color is a visual phenomenon. As a society, manifestos by painters that point a way to new we need to sharpen our skills at visual thinking, understandings. The literature of color, or of at reasoning about what we see in an intelligent popular beliefs about color, also includes the manner. Artists are trained to think in visual writings of Ludwig Wittgenstein, Dante, T.S. terms, but the skill is too important to be taught Eliot, and virtually any author who often refers just to artists. In this post-humanist age, we to color or shares ideas about colors in the need to become seriously interested in course of discussing another subject. understanding what we see, an endeavor more Nobody looks at color with what used to be noble, necessary, and interesting than called the innocent eye. To look without understanding who we are. preconceptions is impossible. Color is In this book I attempt to make sense of something we see, but we adjust our thoughts familiar theories about color. I originally meant about it to conform to traditional and very to show only that discarding worn, meaningless, ancient beliefs passed from one generation to or literary ideas could lay a foundation for the next. We recognize these popular ideas newer, better ideas that were genuinely visual. I about color, woven into the reasoning of the think I have gone farther and suggested that theorists as into our own. But familiar ideas are visual thinking is a necessity, not a nicety. not necessarily good ideas, about to bear

PART ONE Color and Language

If one says "Red" (the name of a color) and there are 50 people listening, it can be expected that there will be 50 in their minds. And one can be sure that all these reds will be very different.

Josef Albers, Interaction of Color

CHAPTER 1 Learning to Use Color Names

I know what colour is because I know red when I see it; I know what red is . . . . Of course the colour red is not the word "red." R. Rhees, "Can There Be A Private Language?"

ristotle rejected an argument by com- face value. Even people who say they know paring it to the misguided reasoning nothing about color know that colors are not of a "a man born blind arguing about sounds or smells. A normal adult usually can provide a colors" (Physics 2.1.11).Aristotle's simile draws general name (for example, red, green, or its authority from the common knowledge that brown) for any color. He or she recognizes color is a visual experience. Persons who have finer variations, to which more esoteric cog- never been able to see have access to color nomens are correctly or incorrectly applied phenomena only indirectly. In theory, those not (, , ). We assume born blind can speak with authority: they could anyone understands that no two colors look learn about colors because they could see them. alike. Therefore, color coding is used on Whether the average person believes he or household appliances and industrial machin- she has learned much about color is another ery. Instructions explain this coding by telling matter. Individuals who say they understand the what to expect if a red light blinks and a green mysteries of, say, computers or macrobiotic light does not. Or conditions are identified cooking probably outnumber those willing to under which the blue lever ought to be make the same claim about color. But pressed. Users are not expected to ask how to the ubiquitous disclaimers cannot be taken at distinguish blue levers from levers.

4 Learning to Use Color Names

Everyone is supposed to know how to do that. realize that objects have names and that certain The aberrants, if any, who do not know, relationships are explained in terms of cannot depend on equipment manuals for help. negatives. Although an adult points to objects in Written descriptions cannot sufficiently clarify showing colors, the names of colors should not the differences people see between yellow and be confused with those of colored objects. A blue. If these perceptual differences could be blue chair is not correctly called a blue. Nor is translated into words without loss of anything blueness a physical part of the chair in the same essential, Aristotle's hypothetical blind person, sense as one of its legs. and all others blind from birth, would have no A child who progresses to this point is difficulty understanding how blue levers differ prepared to acquire competence in use of the from yellow ones. small group of commonplace words regarded as People know more about color than they color names proper. The greater number of realize. How is the information acquired? As these names are monosyllabic: red, , Bertrand Russell pointed out, human beings are yellow, green, blue, violet or , brown, taught to understand color names (a skill , , gray, tan, , and so forth. different from that of distinguishing colors) might be included for its currency in through ostensive definition, the process of Great Britain, although less often used in the teaching the meaning of a word without United States. These words are the primal color explaining it in terms of other words.' To teach names, or simply primals, the small group of color names, the learner is shown colors. An descriptive terms with which everyone is adult can point out blue objects to a child while familiar. repeating the word "blue." The child will The word white (like the word night) is eventually understand that the word refers to a evidently of ancient etymological origin and class of related colors: the aggregate of all cannot be traced to earlier words meaning colors that might be called blue. More blues anything different. Among the remaining exist than those of the objects the adult uses as primals, about half are derived from the names examples. If an intelligent child asks how many, of objects characterized by distinctive colors. the adult has no answer and may change the But yellow, brown, blue, black, and gray show subject. complex etymological relationships. Ostensive definition cannot be the most Yellow, akin to a Sanskrit word meaning primitive means through which children learn. both yellowish and glowing, is also akin to an It requires prior familiarity with at least five Old Irish word for white. Brown is akin to Old ideas. First, although perception is private, High Gothic brun, which means both brown communicating about it is possible. Second, and shining and is used in this manner in communicating differs from perceiving. Third, Beowulf. Blue and black are both related to language is one of several tools for bael (OE, fire). Blue is akin to a Latin word for communicating. Fourth, pointing (as in pointing yellow, and black to a Latin word for burning to a color) can have a purpose. Fifth, people and a Sanskrit word for radiance.' Other primal attribute importance to purposes. color names derive from the names of objects: The child must additionally understand that red, green, violet, pink, tan, purple, orange, objects have names, the circumstance that indigo. links the word chair with certain masses, tac- Red is from Sanskrit rudhira, blood. Green tilely apprehensible aggregates of corporeality (Old English grene) derives from growan, to in the external world. To be capable of osten- grow, a reference to green growing plants. sive learning of color names, the child must Violet and pink are from names of flowers. Tan Learning to Use Color Names 5

is from tannum (Middle Latin), the bark of oak shocks or startles. similarly refers trees crushed to make tannin. Purple is from to an effect that somebody assumed was purpura, Latin for purple fish or sea creature, a electrifying to observers. reference to the mollusk (Purpura lapillus) Like about half of the primals, the largest whose secretion was used to make Tyrian pur- group of nonprimal color names consists of ple dye. Orange is from the Sanskrit naranga, terms derived from the names of objects, such for the fruit. Indigo comes from Greek indikon, as and avocado. Color names drawn from literally, Indian, a reference to dye made in object names are abbreviated similes, and India from the Indigo plant (Indigofera describing color through simile is the tinctoria). syntactical norm in some non-Indo-European Color names borrowed from other lan- languages. In the type of compression common guages may be used differently than in the in English, lengthy concatenations collapse to original tongue. , from the Latin name succinct alternates: "a color like that of for dark blue, is used in English to identify " becomes burgundy or blues that are greenish or (and not burgundy red. When an object name inspires a necessarily dark). Among synonyms for red, color name, the definition of the color name is a and can be traced to an Ara- simile. Burgundy is correctly defined as a color bic word for crimson, thence to a Sanskrit term like that of Burgundy wine. meaning produced by a worm, apparently a Colors are usually named after objects reference to a natural coloring matter. regarded as good examples of that color (figure Cochineal, a color name with an etymologi- 1-1). Occasionally the connection seems cally similar derivation, refers to a dye made remote, in English as in other languages. W. H. from the dried body of an insect (Dactylopius R. Rivers, studying the Mabuiag of New coccus) that lives on cacti in Central America. Guinea, found "a great tendency to invent Errors in use of the primal color names (for names for special colors." But he puzzled over example, misidentifying green by calling it why one native coined a name for a bright blue white) are exceedingly rare. Because of their by comparing it to the muddied brevity, these names lend themselves to from washing mangrove roots (Segall, forming compounds for describing Campbell, and Herskovits 1966, 42). This may intermediary colors. The compounds should be no more mystifying than why the English not be overly long. Bluish green, along with color name is borrowed from the such variants as blue green or blue-green, has a battle of Magenta (1859), at which the commonly understood meaning. Pinkish bluish Austrians were put to rout by the French and brownish purplish white communicates less the Sardinians. effectively, even as a name for a color mixed To understand color names taken from from pink, blue, brown, purple, and white. object names, the prospective learner must A human being might live out his or her life master a further layer of syntactical con- making use of no color names beyond the vention. Colors are not objects, and color primals. Dictionaries of color names illustrate, names are not the names of objects. But object however, that thousands of others exist, often names can acquire a secondary meaning as the used for subtle intermediary colors. One tiny names of colors. The color of a chair can be subset of nonprimal names identifies colors in called olive, although a chair is not an olive. terms of subjective effects they might have on The subtlety of this proposition is beyond the viewers. Shocking pink is incomprehensible grasp of a child too immature to realize that a unless the term means that this variety of pink single word can have more than one meaning

6 Learning to Use Color Names

Fig. 1-1. Color names derived from object names.

alabaster cafe=au-fait curry sanguineiL almond canteloupe delphinium lead pearl caramel ebony ' 3 amethyst cardinal ecru3 persimmon shrimp carrot lily pewter sienna cerises emerald pimento aureolinl chalk fawn lobster pink slate avocado charcoal flame madders pistachio smoke azureZ flamingo magenta pitch banana cherry flesh mahogany strawberry beige3 poppy sulfur bisque chocolate gamboge6 porcelain tangerine blackberry chutney garnet primrose taupe'4 blueberry cinnamon geranium melon pucet° bone mint pumpkin terra cotta brandy clay grape mole purpled toast brass coal gunmetal tobacco brick cochineal heather raven tomato cocoa ' oak turquoise buckskin coffee henna oatmeal ' hyacinth olive russet umber15 burgundy orange rust violet buttercup cranberry jade walnut butterscotch jet palomino watermelon camel currant lapis lazuli paprika sand

Notes: (1) from aureole (halo). (2) from Persian word for lapis lazuli. (3) refers to color of unbleached wool. (4) color of leather made from buffalo skins. (5) French word for cherry. (6) from the name Cambodia; refers to coloring matter made from the yellowish resin from a tree (Garcinia hanburyi) native to that locale. (7) plant name. (8) from a plant that produces dye of that color. (9) from French word for chestnut. (10) from French word for flea. (11) from the name of a mollusk from which purple dye was obtained in ancient times. (12) synonym for bloody. (13) from Persian word for a rich cloth. (14) from French word for mole. (15) from the region Umbria.

During the early twentieth century, of,those objects only, or to the object under proposals for simplifying and standardizing certain conditions. names of colors, in part for the convenience of Among miscellaneous examples, ebony is an industrial users, were put forth by Albert H. allusion to the black wood of the ebony tree, not Munsell, Wilhelm Ostwald, and other color its green leaves. But cherry refers to the fruit of theorists. The systems generally eliminate color its tree. Chestnut means the brown of ripe names drawn from the names of objects. Or, chestnuts, not the green of those that are declining to dignify the matter by discussion, immature. Olive refers to green or brown olives, theorists proceeded as if words such as lemon not those that are black. Flame means the red of and olive were not bona fide color names in the a wood fire, not the blue flame of a gas stove. same sense as red and brown were. Yet real Wine usually means the color of red wine. color names are neither more nor less than Tulip is rarely used as a color name; the words people use to name colors. Few flower exists in many colors. Yet rose, in both differences can be identified between the primal English and French, has acquired a widely and nonprimal names, except that more familiar meaning as the color of, so to speak, complicated language conventions need to be rose-colored (or pink) roses, never those that mastered to understand the latter. References to are yellow or white. Violet, similarly, is the objects are often references to certain parts

Learning to Use Color Names 7 color of purple violets, although the flower can Umbria or the etymological linkage between also be blue, yellow, or white. the place name and the color name. Similarly, Flowers provide some of the more esoteric those who know nothing about the art of examples of unusual associations. Madder, a China, Pompeii, or Raphael can learn the con- color name used for paints (rose madder, mad- ventions for use of Chinese red, Pompeiian der lake), is not related to the yellow flowers red, and Raphael blue. of the madder plant. The allusion instead is to The rules by which color names are derived the red dye manufactured from the plant's from object names are complex and intersect root. Saffron is not the color of the saffron with rules for the use of adjectives and nouns. flower, a purple crocus. The word refers to the The English language includes two groups of yellow-orange of the dried stigmas of the compound words that couple the name of a flower. color with that of an object. Sequence fun- Color names borrowed from object names ctions as a determinant. If the color name fol- need not always allude to the color of the lows the object name, the compound is a color named object. In a subclass, the object cited name (, grass green, midnight black, is a geographical entity: a city, district, or , lemon yellow). Although excep- nation associated with a particular color or col- tions can be cited (see below), usually no com- ored . Although sienna and Naples pound is a color name if the order is reversed, as yellow are color names derived from names of in redbird, bluebird, blackbird, whitefish, artists' , each of the pigments is goldfish, and bluegrass. named after a place (respectively, Siena, Sky blue is a color name, although blue sky Naples) once noted for its manufacture. is not. Sky blue is not synonymous with blue; A variant linkage between color name and it names a narrower range. Grass green is the geographical name turns on allusion to artifacts name of a type of green, although bluegrass is and works of art from the indicated locale. Chi- the name of a type of grass. Within the frame- nese red points to the reds in Chinese lacquer work of existing convention, bluegrass, work; Pompeiian red, to the red backgrounds although a type of grass, could reasonably of wall paintings at the Villa of the Mysteries, acquire an additional meaning as the name of the Villa Boscoreale, and a few other Roman a color resembling that of bluegrass. Red lead, houses at Pompeii. In an alternate that cites the identified as a color name in color-name dic- name of an artist rather than that of a place tionaries, was evidently the name of a sub- where works of art were created, Raphael blue stance (red oxide of lead, Pb304) before being is the color of Mary's robe in several of adapted as the name of a color. Raphael's paintings of the Madonna and Child. Although colors are often named after Color names that allude to works of art bear a objects, objects are rarely named after colors time stamp, because they cannot be of earlier Rouge (the cosmetic) is from the French word vintage than the art to which they refer. The for red, a borrowing that has acquired new color called Raphael blue must have been meaning in passing from one language to known by some other name to Raphael's another. In the dubious argument of the predecessors. painter and color theorist Moses Harris Human beings can learn to use a word in (1731-85), "the word orange seems indeed as if a conventional manner without knowing its the colour took its name from the fruit, but derivation. Umber is understandable as a name the fruit took its name from the colour, for the for dark brown, even for those uninformed proper name of the fruit is the orange citron" about either the central Italian district of (Harris [1766] 1963, 8). If orange is a contrac- 8 Learning to Use Color Names

tion of orange citron, then strictly speaking the Figure 1-2. Color coding used to indicate the ohmic value of carbon resistors (resistance given in ohms). fruit is not named after the color. And orange is Four colored bands encircle the resistor. The two generally fruit before color. Naranja (Span-ish) near the end indicate the first two digits of the ohmic value. The third indicates the number of zeros to add seems to allude to the shape (rather than color) to these digits. The fourth indicates the tolerance. of oranges, because the word can also mean a (After Donald E. Herrington, How to Read Schematic Diagrams (New York: Bobbs Merrill, 1962) p. 23.) cannonball of the same size as an orange. People, or groups of them, are named after Color Digit Multiplier Tolerance colors more often than are other objects. (+ or -) 20.00% Examples, however, are infrequent. The name Black 01.00 1.00% of the Hindu deity Krishna comes from the Brown 1 10.00 2.00% Red 2 100.00 3.00% Sanskrit word for black. In English, Black, Orange 3 1000.00 GMV Brown, Green, Gray, and White, along with Yellow 4 10000.00 5.00% Green 5 100000.00 6.00% such variant spellings as Browne, Greene, and Blue 6 1000000.00 12.50% , are commonplace family names. Pink, Violet 7 10000000.00 30.00% Gray 8 .011 10.00% yellow, purple, and orange (although Holland White 9 .101 5.00% has its House of Orange) are almost never used Gold .102 10.00% Silver .012 20.00% for this purpose. Red, a common nickname for No color those with auburn hair, is rarely a given name or surname. the ohmic value of carbon resistors (figure 1-2). Among groups larger than family units, race Because colors have no correlation with num- (described by anthropologists as if it were an bers or ohms, the coding could easily have extension of family, clan, or tribe) provides the been reversed. most striking example in which the names of Theories about the origin of color names, objects (of groups of humans) are borrowed like those about the origin of mathematics, are from those of colors. Like the practice of schematic and conjectural. The most familiar adopting color names for family names, that proposes two historical stages in the develop- of imagining humanity as a whole subdivided ment of color vocabularies. Among primitive into black, brown, red, white, and yellow race peoples, the color of an object is said to be exists on a nebulous borderline between nam- described by comparing it to that of another ing and coding. In medicine, a few patholo- object of similar color, adopted as the implied gies named after colors refer to abnormal norm. Green items, say, are identified as the seen in the human body in those conditions, color of (green) leaves. In a later development, usually in the skin. Among examples, jaundice human beings coin color names that are comes from the French word for yellow, and abstract, that do not refer to objects. cyanosis, which describes the blueness of As Segall, Campbell, and Herskovits have "blue babies," is from kyanosis, a Greek word pointed out (1966, 40), for Grant Allen and for dark blue. other nineteenth-century theorists "color Although rarely named after colors, objects terms develop first where color distinguishes are often either coded or symbolized by means among objects that are otherwise similar. of colors or color names. The usage can be Where colors and objects go together uni- arbitrary, in the sense that the color need not formly, the object name suffices. The avail- reveal anything about the object. For example, ability of pigments and dyes facilitates the red is assigned a value of 2 and white has a development of abstract color words that are value of 9 in the colored bands used to code applicable to the color no matter upon what Learning to Use Color Names 9

object it is found. Color terms are initially this, nineteenth-century analyses are question- metaphorical extensions of what are originally able on two points. First, it cannot be true that object names, or else of pigment and dye color names originally developed to permit a names. distinction to be made between objects alike Variations of these theories persist, because in form but different in color. If that were the identifying colors by simile has been shown case, there would never have been a need to to be common among primitive peoples. That identify grass as green: no item exists that looks the practice itself is primitive is unlikely. In the exactly like grass except that it happens to be English language, our largest group of color purple, red, orange, or blue. names consists of condensed or abbreviated Second, the etymological record does not similes. Describing a color as "like (green) support the conjecture that the earliest color leaves" is not significantly different from iden- names referred to objects while more modern tifying the colors of objects by those words names are abstract. A better argument can be listed in figure 1-1. Each names a color but also made that the circumstances are the reverse. (because simile is implied) an object available As noted earlier, the color names blue, yellow, as an example of the color. Use of simile is white, black, and brown are of such ancient more than just commonplace in naming colors etymological origin that they cannot be traced in English. It remains the standard means of to earlier words that have any other meaning. inventing color names, as ubiquitous as jargon, Later color names, which unvaryingly point to as irrepressible as graffiti. objects, are abstract only in the limited sense Many of our comparisons are of recent vin- that they can be shown to be extrapolations tage.Electric blue, as a color name, cannot pre- from (or abstractions based on) the names of date the term electricity. Fire engine red, objects. tomato red, and tobacco brown are similarly Early theorists might have considered such time stamped. So is pistachio, which refers to color names as red, green, and uncon- the color of pistachio ice cream, not pistachio nected to objects, hence abstract. But the dis- nuts. junction is nominal. Each of these words Avocado became popular as a color name, comes from the name of an object, though in particularly among American manufacturers of a language other than English. Because bor- kitchen appliances, during the period when rowing color names from object names is so this tropical fruit acquired wide commercial common, uncertainty surrounds even blue, distribution. Avocado is not used as a color yellow, white, black, gray, and brown. name for office furniture, and innovative color Although these six names cannot be shown to names are more common in industries in be derivatives from object names, this is which changing fashion is a factor. Writers of insufficient evidence that they are not. Their advertising copy for department stores can be origins may be lost in antiquity. expected to tantalize prospective customers in During the past 150 years, advances in pig- coming decades with such stylish offerings as ment and dye technology set a course for mod- bath towels of intergalactic blue and under- ern art and revolutionized public taste. The wear of astronaut green. Whatever criticism catalyst was an English teenager, William might be made about such coinages, people Henry Perkin (1838-1907), who in 1856 syn- will understand what they mean. thesized mauve from coal tar, evidently in an Use of simile in naming colors is not inspired blunder. The dye-making technology unusual in English. The form is sophisticated based on Perkin's discovery swept the world. for its purpose, rather than primitive. Beyond Even before the first exhibition of Impres- 10 Learning to Use Color Names

sionist painting, Navaho rug weavers were their similarities and differences" (Russell using the aniline dyes synthesized in chemical 1948, 69). If vermilion is called a fiery red, the laboratories. Traditional vegetable dyes were description, although ambiguous, conveys put aside because their colors were less bright. some information to anyone who has seen red. The commercial success of aniline dyes Theoretically, the learning process could pro- may have led Grant Allen, whose Colour-Sense ceed in reverse order, with common names was published in 1879, to overestimate both taught by reference to those that are unusual. the historic importance and the modernity of If a person is familiar with vermilion, red can dyeing and coloring processes. From the be described as either a color of the same type Paleolithic painters onward, every human soci- as vermilion but not always so fiery, or the ety known has used staining or coloring color class to which vermilion belongs. materials. We cannot reasonably ask whether Sensibility to color depends on vision a society without pigments or dyes would be before vocabulary. Recognizing a color with- more likely to identify colors by pointing to out knowing a particular, rather than generic, objects. The question is whether such a soci- name for it is a common, annoying experience. ety exists or has ever existed. It causes little practical difficulty. Magenta Staining and coloring with found or pieces of paper can be sorted from vermilion manufactured materials is a primitive impulse ones, even by persons who never knew, or in the life of the individual, as in that of the cannot remember, the words magenta and ver- human race. Smearing of feces by children, milion. Awareness that vermilion is a type of investigated in the psychoanalytic literature, red is not enough to enable a person to deter- provides an early experience in changing the mine which of several red color chips is ver- color of a surface by applying colored mate- milion. And an observer who believes that a rial to it. It also provides experience in use of certain object is vermilion may find that not the human (or primate) hand to manipulate everyone agrees. The red in question, others objects. The attention of modern children is may say, is not fiery enough to be called ver- redirected to wax crayons and coloring books. milion. Paper is provided, to which children-but also A child requires considerable sophistication chimpanzees-enjoy applying paint. A later to understand these everyday mysteries. Color improvement of motor skills allows children names are applied according to visual criteria. to be taught writing, a specialized form of But people often disagree about application of applying color to a surface. the criteria. The child, if able to grasp this final Bertrand Russell suggested that the more subtlety, will have come a long way since the ordinary color names, such as red, can only day someone pointed to a chair and pro- be learned ostensively, but "the less common nounced the word "blue." ones, such as vermilion, may be described by CHAPTER 2 Color as a Continuum

Looked at from the point of view of an individual thinker, the act of naming is the first step in knowledge. At the very beginning of modern logic, Thomas Hobbes rightly said: "Reason is attained by industry, first in apt imposing of names." Daniel Sommer Robinson, The Principles of Reasoning

olor is a continuum because it forms be completed, but how anyone could find a way the fabric of visual homogeneousness: to begin. of the uninterrupted expanse of what In practice, the fineness of the intermediary I see. In a more immediate sense, color is a ranges in the color continuum is limited by continuum because any two colors are sepa- human perception. The constraint is rated by a range of intermediary colors (figure acknowledged in those colorimetric studies that 2-1). The number of these intermediaries may propose the minimum unit of a justnoticeable be infinite if the color continuum is analogous difference (just noticeable to human beings) to the number continuum and we follow the between two nearly alike colors. The possibility reasoning of the mathematician Georg Cantor cannot be ruled out that a race with more (1845-1918). Because the number of fractions exquisite visual acuity would see differences between any two consecutive integers is between two reds that all human beings believe infinite, Cantor's work implies that the num- are alike. ber series has neither a beginning nor an end. Precisely because a just-noticeable differ- For anyone who imagines making a tally of ence is relative to an observer, people vary in every color variation, the model is serendi- ability to discriminate within the color con- pitously apt. The question is not whether the job tinuum. Some individuals are better than of counting all the colors in the world could others at deciding whether colors are identi-

12 Color as a Continuum

Figure 2-1. Color as a continuum. The Munsell solid Names and the Continuousness arranges colors according to the parameters of hue, value, and chroma. of Color Naming the other living creatures is the first Green Extremely Greenish Blue-green task Adam performed in the Garden of Eden. Very Greenish Blue-green Although presented in the biblical story (Gen. Greenish Blue-green Blue Green 2:19) as a gesture of dominion, the more Green blue immediate purpose of naming is to facilitate Bluish Green-Blue Very Bluish Green-Blue talking about the entities to which names have Extremely Bluish Green-Blue been given. The Bible offers no clue about how Blue Adam decided which names to use. His descendants apply names according to what is perceived as discrete. Horse and cow are different names because a horse is not a cow. cal. The skill is partly innate, partly acquired, Nor does one animal blend into the other, partly dependent on temperamental factors: on except in the mystical sense that the universe whether the observer has the confidence to can be regarded as a unified whole. make a careful evaluation and to express a The key problem in naming members of a considered judgment. continuum is that there may be no members The ability to recognize that red is not blue, that can be separated from the whole. Nor virtually universal among those of normal does language offer a mechanism that allows vision and comprehension, implies the further names to blend into one another, as the understanding that differences among colors can individual colors blend in the color con- be regarded as quantitative. Relying on a tinuum. The familiar expedient is to pretend, mathematical metaphor that may not be to defer to utility by talking as if we could appropriate, can impress an divide the unity of a continuum into observer as either great or small. Some colors multiplicities. match, others almost do, still others bear little If a society is to develop a workable system resemblance to one another. of color names, the arbitrary points at which Among many similarities between the color the color continuum is to be "cut" (or imagined continuum and the number continuum, the as segmented) must be determined as exactly name of each is a coinage of convenience for as possible. To avoid tempting everyone to an entity we cannot locate. Like the idea of form his or her own opinion about whether number as a continuum, the idea of the many blue and red are different colors, the shades of color organized in a hierarchy is a determination is best sustained through mental construction. It develops from the consensus. Consensus is also required on the everyday recognition that an unknown num- rules for matching segments with names. ber of colors lies chromatically "between" Should blue be called blue? Or should it be other colors, an encounter with that which called yellow? exceeds measure. Just as the number series Everyday experience teaches that these continues forever and the sky has no discern- tasks are difficult to accomplish, or we quickly ible edges, the number of shades of bluish arrive at an impasse from which logic cannot green between blue and green is indeter- extricate us. Two observers may agree that one minable. color is blue and another is green. Even begin-

Color as a Continuum 13 ning from, so to speak, the same premises, prehensive information about objects. Carnap they may not agree on the name for a third mentioned, as an example, a black-and-white color "between" the first two. What one photograph of city buildings, incomplete in observer calls bluish green, another may insist failing to record the colors of the buildings. is greenish blue. Color names illustrate the more successful type Absence of agreement reflects uncertainty of language, conveying complete information. about naming, rather than about sorting colors. If you hear the word blue, Carnap pointed out, Viewers who differ about whether to call a "you immediately imagine blue" (Carnap color greenish blue or bluish green usually will 1966, 114). not disagree about how to arrange blue-green Carnap's argument raises the question of color chips in a row according to relative whether blue is a clearly defined entity. Can degree of blueness or greenness. What can be we visualize the color by, in a manner of speak- seen in looking at color has a simplicity with- ing, translating it into a mental picture? I pre- out correspondence in language. We under- fer more limited assumptions. On hearing the stand what we see, but search for words to word blue, the listener understands that the explain what is understood. term is a generic name for a class of colors, the Whether a shade of color is greenish blue class containing many individual shades of (by implication, a type of blue),or bluish green blue. In Russell's reminder of the truism, these (a type of green) cannot be settled by deter- "many shades [have] different names; there is mining the point at which the blue segment , aquamarine, peacock blue, and so of the color continuum ends and the green on" (Russell 1948, 126). begins. Beginnings and endings within the Because the number of is continuum are located where people want to indeterminable, the question is how to imagine place them. A more subtle constraint is that all of them at the same time, an existentially because temporality (the passing of time) is absurd endeavor. Relying on the familiar man- implied, the color continuum has no point at ner in which dictionaries define blue, a set of which, in any physical or experiential sense, instructions for imagining every blue might color stops belonging to the class blue and advise imagining one portion of the spectral begins belonging to the class green. continuum or hue continuum: the range of Language consists of a series of little boxes colors in the running from those blues called words. Color is not a series of little that are most green to those that are most boxes. Language and color match up poorly, purple. which is why color names are usually The simplicity of the prescription is regarded as ambiguous, often frustrating to treacherous. Every blue includes a greater use and difficult to understand. Among a few number than those of the solar spectrum, dissenting voices, the philosopher Rudolph because the spectral continuum fails to span all Carnap contended that color names are not dimensions of the color continuum. Varia- difficult to use, communicate with more than tion can be found among blues beyond usual effectiveness, and function as reason- whether they tend toward green or toward ably unequivocal descriptive devices. purple, their immediate spectral neighbors. Carnap argued that language is an abstrac- Blues can vary according to whether light or tion and that two types of language exist. The dark, bright or muted, matte or shiny. first, an incomplete code, cannot convey com- The developed by the American 14 Color as a Continuum

painter and educator Albert H. Munsell is a one another (to expedite seeing all of them at three-dimensional construction (other color once). solids have come from other theorists). Its pur- Imagining cannot rise to the task because pose is to provide a model for relationships of the contradiction in terms. There is no such within the color continuum (figure 2-2). The thing as a visual image of the concealed Munsell solid arrays color samples according interior of an opaque solid (here, a color solid). to hue, value, and chroma. Hue is "that qual- Nobody could ever see such a thing or know ity by which we distinguish one color family what it would look like. The visual or visualiz- from another, as red from yellow, or green able has limitations, which is why we cannot from blue or purple." Value is "that quality by imagine how nothing would look if it could which we distinguish a light color from a dark be seen. one. Color values are loosely called tints and Proposing that we "imagine blue," Carnap shades, but these terms are frequently misap- refrained from conjecturing about what pic- plied. A tint should be a light Value, and a tures might come to mind on hearing the word shade a dark Value, but the word shade has color. To direct an artist to take a canvas and become a general term for any type of color color it "color" is meaningless. The task can- so that a shade of yellow may prove to be not be performed because, as the name of a lighter than a tint of blue." Chroma refers to genus or class, color has no direct visual or pic- the strength of a color, "that quality of color torial equivalent. by which we distinguish a strong color from The color that corresponds to blue is as elu- a weak one; the degree of departure of a color sive as that corresponding to color, for simi- sensation from that of white or gray; the inten- lar reason. Blue is generic. It names a range of sity of a distinctive Hue; color intensity" (Mu- colors rather than an individual shade we can nsell [1905] 1961, 15-16). isolate with the eye or in the mind's eye. By Like color atlases, which array graded color encompassing every member of its class, blue swatches in grids on multiple pages, the color fails to point to any particular blue. We can- solid implies a three-dimensional color con- not direct an artist to take a canvas and paint tinuum. Dimensions beyond three may exist, it (generic) blue, let alone with the further although we do not know how to incorporate stipulations that the desired color be every a fourth or further spatial dimension into the blue but no particular blue. No object can be model graphically. Colors, for example, can every member of its class yet at the same time vary in shininess of surface. The Munsell sys- no particular member. tem acknowledges this parameter but suggests The hypothetical artist might, at least the- no way to incorporate it into the three- oretically, divide the canvas into compart- dimensional color solid. ments, color each a different shade of blue, Despite the preferences of compilers of dic- and thereby include every blue. But every blue tionaries, any three-dimensional color solid is (which is what blue means) has no visual superior to the hue continuum as a model of equivalent unless we can be certain about the color continuum. The challenge for which colors are unambiguously blue. The prospective imaginers is to visualize the array greenish blues (which are blue) would have to of shades of blue as they appear in a model of be separated from the bluish (which are this type. The blues lie on an infinity of col- green). The purplish blues would have to be ored planes arranged on three axes. We must separated from the bluish , the whitish picture the planes behind one another (as blues from the bluish , and so forth. arranged in color solids) but also not behind The task cannot be accomplished. Color as a Continuum 15 sensus would be required, and no consensus tute generic terms for proper names as a form of of any firmness is available. For this reason, it address. is beyond human capability to speak authorita- Among generic or class names people use tively to the question of how many colors, and when addressing one another, the more for- which, are properly included under the rubric mal are called titles: Doctor, Professor, every blue. The term is experientially meaning- Madame President. Syntactically similar alter- less, a convenient fiction. The artist dividing nates are less socially acceptable. They range the canvas into compartments, each to be col- from the marginally discourteous ("Mister !") ored a different blue, would have no way of to the overtly hostile (racial, ethnic, and gen- determining how many compartments to der slurs). The class name, used as a surrogate provide. for the proper name, is understood not to be I doubt we imagine blue according to any the name of the individual being addressed. definition of that generic color that is more We all know that nobody's name is Madame than nominal. Another way of understanding President or Mister. the trains of thought that occur when people Names for individual shades of color are hear color names is to assume that the mind similarly acknowledged to differ from the wends its way into symbolism, the game of generic name of the class. The individual ques- pursuing an elusive object by appropriating tioned on why any and every blue is being another object as its surrogate. Thus, a called "blue" can reasonably be expected to hypothetical listener hears "blue," a word confess awareness that specific names exist for recognized as the name of a color class. The particular blues. The likelihood is that the class includes, as Russell has indicated, a large, namer cannot recall them, is uncertain of how but unknown, number of varieties of blue. The to apply them, or for some reason lacks listener seeks a single blue to imagine (the interest in offering them. Everyone knows, or thrust is toward specificity, not generalization), appears to agree, that words such as blue are relying on context for clues. If the admiral generic, apply to broad classes of color, and wore a blue uniform and the sky was blue, the are not the names of the individual members same blue cannot be meant in both cases. If of their respective classes. the listener finds no pointer toward finer cat- The expediency of using the class name for egories, he or she selects a single shade of blue the individual member, with the proviso that that seems pure or typical, thus appropriate as it is understood to be "not really the proper a representative for the class. The barriers to name," sheds light on the survival, in English, imagining this single shade are less formidable of the enormous number of color names col- than those to imagining blue in general. lected in specialized dictionaries. Some of the A counterargument rests on the truism that names are rarely used, and many people pro- a word can have more than one meaning. Blue fess uncertainty about the meaning of others. can name a broad class of colors but also par- Is peacock blue equivalent to dark cerulean? ticular members of the class. Perhaps imagin- Does cerulean differ from turquoise? A large ing any single shade is imagining blue, though repertory of infrequently used color names is in the particularized sense of the word. The easy to dismiss as superfluous, a type of lin- appeal of the counterargument is that many guistic appendix with no identifiable utility. people, unconcerned with fine distinction, I prefer to believe the body of names serves habitually use the name blue for any variety a purpose, because we realize it exists. By of the color. The usage conforms to conven- existing (or by being known to exist), it tions similar to those that allow us to substi- assumes the potential for satisfying the uni- 16 Color as a Continuum

quely human requirement that every object of Names, including color names, can be oper- interest to human beings have a name. No par- ationally defined as a recognition of the dis- allel requirement exists that everyone know or crete: an acknowledgment that an entity exists use all the names. My not knowing the name of apart from or (more likely) in imagination can every human being or of every species of be extrapolated from the unity of everything animal does not interfere with my conviction in the universe other than itself. that it is appropriate for each to have one.

CHAPTER 3 Understanding Color Names

And how important it is to know how to mix on the palette those colours which have no name and yet are the real foundation of everything.

Vincent van Gogh, Complete Letters

arnap's argument in favor of the suffi- in all American flags is similar, manufacturers ciency of color names takes issue with must agree on what blue means in this case. the proverbial wisdom that any pic- The agreement, which cannot be verbal, is for- malized through the production of swatches: ture is worth a thousand words. In opposing pieces of paper painted or printed in the color blue and a photograph of buildings, he tells us to be standardized. Visual comparison a single word communicates more than a pic- between swatch and object tests whether the ture. A picture of buildings, whether mono- blue of flags on the assembly line matches the chrome or colored, communicates more necessary shade. The deaf can make this com- information than the word buildings. A blue parison as easily as can anyone else, even those color swatch communicates information the born profoundly deaf who never have heard color name does not convey. The swatch anyone pronounce "blue." shows definitively, as words cannot, which Visual comparisons are more accurate than variety of blue is meant. those made by instruments (Evans 1948, 203) When more than one person must interpret and facilitate color matching more effectively a color name, words alone cannot identify than any attempt to interpret statements that which color is intended. Color standardizing include color names. Consider the commercial addresses problems that arise in industry as a requirements that vitamin C pills not be too result. To ensure, for instance, that the blue

18 Understanding Color Names

yellow, or that gray file cabinets rolling off an ence to those that are common. Because there assembly line match the color of those are tens of thousands of , teaching the manufactured last year. Accomplishing these difference between rufous and fuscous can goals requires the use of swatches, because only be accomplished by showing the colors. words cannot communicate sufficiently. How Words do not convey enough information, or yellowish is too yellow? Showing, rather than do not convey the right information. Osten- telling, avoids misunderstanding. sive definition is more than a technique by In the manufacture of American flags, the which children can be taught a basic vocabu- assumption can be made that thirteen will lary of color names. It is also the only adequate mean the same thing to different persons at means of making fine discriminations between different times and places, but blue will not. colors and communicating them among peo- We do not need standards to show how thir- ple. Color names are useful to the extent that teen (as in thirteen stripes) should be inter- they relieve human beings of the nuisance of preted. But we need them for blue. The carrying packages of color samples with them codability of color names-their adequacy in wherever they go. conveying information about that to which Dictionaries of color names have been com- they refer-is low compared to that of names piled and reveal a large repertory. Robert Ridg- and notations for numbers. way, curator of the Department of Birds of the Efforts to achieve greater specificity, al- United States Museum, published a color dic- though of practical value, do not eliminate the tionary in 1886 and an illustrated edition in problem. Colors properly called dark cerulean 1912. It includes 1,113 painted color samples blue are not as numerous as those properly identified by name. Munsell notations have called blue. Dark cerulean blue, nonetheless, been published for the samples. The Maerz and has a large number of subsets or varieties. Paul Dictionary of Color includes 4,000 color Hundreds or thousands of colors are accepta- names keyed to 7,000 samples. The National bly identified as dark cerulean blue, but the Bureau of Standards has published The Inter- colors fail to match one another. Society Color Council Dictionary, and Euro- The codability of some color names is so pean dictionaries have come from the British low we cannot be certain they communicate Colour Council, the Royal Horticultural Soci- (or cause the listener to imagine) anything. Few ety, and the Societe Francaise des Chrysan- people conjure up a vivid mental picture on themistes. hearing "fuscous." The additional information Many of the names in color-name diction- that fuscous is a standard color name used in aries are not widely familiar, used in a consis- ornithology would not be very helpful. Fus- tent manner, or often employed. If language cous is a type of brown, displayed in a sample attains highest utility when used with greatest swatch in Frank M. Chapman's Handbook of economy (when every object has a name of its Birds of Eastern North America. In describing own and no object has two names), each name the colors of the feathers of birds, ornitholo- ought to apply to only one range of color. Real- gists must not confuse fuscous with rufous, ity falls short of this ideal. The twenty-four another variety of brown. color names in the list below, all referring to The limited usefulness in knowing that fus- varieties of red orange, are not synonyms; they cous and rufous are both brown suggests a identify more than one range of color. Nor are need to qualify Russell's assertion that children they names for twenty-four different ranges; can be taught unusual color names by refer- overlaps occur in some cases. Without color

Understanding Color Names 19

samples, nobody would be able to explain what he or Society Color Council Dictionary lists she regarded as the proper use of each name. sixtynine for the range of color standardized as moderate reddish orange (ISCC-NBS color no. 37). More than a hundred are given for others vermilion Japanese red (figure 3-I). carnelian Spanish red Inter-Society Color Council compilations persimmon Naples red reveal that vermilion is used, but not flame red Mars red consistently, for the color range standardized as red lead scarlet moderate reddish orange (color no. 37). Egyptian red cochineal Dissenters reserve vermilion for strong reddish Pompeiian red turkey red orange (color no. 35), a more intense orange. Chinese red red earth The lack of consensus suggests data has been crimson Morocco red collected from members of the general public, poppy Venetian red who are often uncertain about how to use lacquer red unusual color names. But the information in this madder red English red case comes from color name dictionaries; the Although the twenty-four names for var- lack of agreement is that of authorities. In ieties of orange-red do not all refer to the same nontechnical dictionaries, lapses in cross- shade, a given shade of color is sometimes referencing of color words are common. The known by dozens of popular names. The Inter- American College Dictionary identifies henna

Figure 3-1. Names for ISCC standard color no. 182 (moderate blue). iources: 1-72, Maerz and Paul (1930); 73-86, Plochere (1948); 87-105, Ridgway (1912); 106-19, Taylor, Knoche and Gran ville (1950); 120-25, Textile Color Card Association (1941).

1. air blue 26. diva blue 51. nikko 76. coronet blue 101. marine blue 2. Antwerp blue 27. Dresden blue 52. orient blue 77. cosmic blue 102. Orient blue 3. Armenian stone 28. Dumont's blue 53. pilot blue 78. cruise blue 103. oxide blue 4. Asmalte 29. Dutch 54. pompadour green 79. deep water 104. 5. blue de Lyons 30. empire blue 55. porcelain 80. Lake Como 105. Vanderpool's 6. blue ashes 31. enamel blue 56. 81. Lake Louise 106. bright cerulean 7. blue aster 32. English blue 57. queen blue 82. Neopolitan night 107. bright navy 8. bluebell 33. eschel blue 58. queen's blue 83. palace blue 108. cerulean blue 9. blue bice 34. flaxflower blue 59. Raphael 84. queen blue 109. Copen blue 10. bluebird 35. gentian 60. rapids 85. Riviera 110. dark blue 11. bluer 36. Harlem blue 61. resolute 86. theatrical blue 111. deep blue 12. blue ultramarine ash 37. Hungarian blue 62. 87. alizarine blue 112. deep cerulean 13. blue verditer 38. Infants 63. Sander's blue 88. Antwerp blue 113. Delft blue 14. Britanny 39. jay blue 64. Saunder's blue 89. Blanc's stone 114. Dells Robbia blu 15. cadet blue 40. king's blue 65. Saxony blue 90. cadet blue 115. Dutch blue 16. cathedral blue 41. Lambert's blue 66. small 91. Chapman's blue 116. lapis lazuli 17. celestial 42. laundry blue 67. smaltino 92. chessylite blue 117. medium blue 18. centre blue 43. lime blue 68. triumph blue 93. China blue 118. sky blue 19. ceramic 44. Limoges 69. Tuileries 94. 119. strong blue 20. chessylite blue 45. Madonna 70. virgin 95. dark cadet blue 120. bluebird 21. China blue 46. mineral blue 71. wireless 96. dusky greenish blue 121. hydrangea blue 22. cobalt glass 47. mosaic blue 72. blue 97. 122. lustre blue 23. commelina blue 48. mountain blue 73. bohemian blue 98. gendarme blue 123. Majolica blue 24. copper blue 49. Murillo 74. ceramic 99. Hortense blue 124. old China 25. Daphne 50. national blue 75. classic blue 100. jay blue 125. Peking blue

20 Understanding Color Names

as a name for a reddish orange dye. An entry Figure 3-2. Color (pigment) names used by three manufacturers of artist's oil paints. for henna as a color name identifies it as Even among manufacturers of artist's oil reddish brown, leaving open the question of paints, pigment names tend to be used loosely. whether henna is the correct name for the color of articles dyed with henna. Name of Paint 1 2 3

Among manufacturers of artist's oil paints, alizarin crimson x x x no two use exactly the same color names or alizarin crimson, golden x alizarin carmine x apply them to the same shades of color (figure aureolin x 3-2). We must look at the paint in the tubes to brown madder x x x burnt umber x x x find out whether Rembrandt's Chinese cadmium orange x x vermilion or its Dutch vermilion is more similar cadmium red x cadmium red lightest x to Grumbacher's vermilion (Chinese). Winsor & cadmium red extra pale x Newton's cadmium yellow pale may or may not cadmium red light x cadmium red medium x closely match Grumbacher's cadmium yel- cadmium red deep x x low light. cadmium red extra deep x cadmium lemon x The National Bureau of Standards, which cadmium yellow light x has an interest in industrial color standards, cadmium yellow pale x cadmium yellow extra pale x sponsored a study of color names in 1932. The cadmium yellow medium x x x immediate inspiration was that "E. N. Gather- cadmium yellow deep x x x cadmium yellow orange x coal, member of the U.S. Pharmacopoeial cadmium green x Revision Committee, protested the selection of cadmium green light x cerulean blue x x color names used to describe chemicals and cerulean x drugs . . . in particular, the term: `blackish chromium oxide green x oxide of chromium x white' " (National Bureau of Standards n.d. b, oxide of chrome mat x 1). The case provides an example of failure to x Cambridge green x understand a compound color name, although Cambridge red x its components were familiar. Cambridge violet x Cambridge yellow x Gathercoal knew, because everyone x knows, what white, black, and the derivative cobalt blue light x x cobalt blue deep x x blackish means. This did not explain blackish cobalt rose x white to his satisfaction. Perhaps he thought cobalt violet light x x cobalt violet deep x x of the colors as opposites, a condition that cobalt violet dark x implies blackish white is a contradiction in cobalt green deep x cobalt green light x x terms. Or, unlike Carnap's hypothetical lis- deep x tener hearing the word blue, Gathercoal might chrome yellow lemon x chrome yellow light x have been unable to imagine anything upon Chinese vermilion x encountering the color name blackish white. vermilion Chinese x Dutch vermilion x As first chairman of the Inter-Society Color Council (ISCC), organized to facilitate 1 = Grumbacher 2 = Winsor & exchange of information about color among Newton industrial and scientific groups, Gathercoal 3 = Rembrandt supervised the preparation of what was to be an improved method of cataloging and nam- ing it to its present form a decade later. Among ing colors. The National Bureau of Standards criteria, Gathercoal aimed for a system "suffi- published the ISCC-NBS system in 1939, revis- ciently commonplace to be understood, in a

Understanding Color Names 21

Figure 3-3. Names of the 267 major color classes in the ISCC-NBS method of designating colors.

1. vivid pink 68. strong orange yellow 135. light yellowish green 202. very pale purplish blue 2. strong pins. 69. deep orange yellow 136. moderate yellowish 203. pale purplish blue 3, deep pink 70. light orange yellow 137, dark yellowish green 204. grayish purplish blue 4. light pink 71. moderate orange yellow 138. very dark yellowish green 205. vivid violet 5. moderate pink 72. dark orange yellow 139. vivid green 206. brilliant violet 6. dark pink 73. pale orange yellow 140. brilliant green 207. strong violet 7. pale pink 74. strong yellowish brown 141. strong green 208. deep violet 8. grayish pink 75. deep yellowish brown 142. deep green 209. very light violet 9. pinkish white 76. light yellowish brown 143. very light green 210. light violet 10. pinkish gray 77. moderate yellowish brown 144. light green 211. moderate violet 11. vivid red 78. dark yellowish brown 145. moderate green 212. dark violet 12. strong red 79. light yellowish brown 146. dark green 213. very pale violet 13. deep red 80. grayish yellowish brown 147. very dark green 214. pale violet 14. very deep red 81. dark grayish yellowish brown 148. very pale green 215. grayish violet 15. moderate red 82. vivid yellow 149. pale green 216. vivid purple 16. dark red 83. brilliant yellow 150. grayish green 217. brilliant purple 17. very dark red 84. strong yellow 151. dark grayish green 218. strong purple 18. light grayish red 85. deep yellow 152. blackish green 219. deep purple 19. grayish red 86. light yellow 153. greenish white 220. very deep purple 20. dark grayish red 87. moderate yellow 154. light greenish gray 221, very light purple 21. blackish red 88. dark yellow 155. greenish gray 222. light purple 22. reddish gray 89, pale yellow 156. dark greenish gray 223. moderate purple 23. dark reddish gray 90. grayish yellow 157. greenish black 224. 24. reddish black 91. dark grayish yellow 158. vivid bluish green 225. very dark purple 25. vivid yellowish pink 92. yellowish white 159. brilliant bluish green 226. very pale purple 26. strong yellowish pink 93. yellowish gray 160. strong bluish green 227. pale purple 27. deep yellowish pink 94. light olive brown 161. deep bluish green 228. grayish purple 28. light yellowish pink 95. moderate olive brown 162. very light bluish green 229. dark grayish purple 29. moderate yellowish pink 96. dark olive brown 163. light bluish green 230. blackish purple 30. dark yellowish pink 97. vivid greenish yellow 164. moderate bluish green 231. purplish white 31. pale yellowish pink 98. brilliant greenish yellow 165. dark bluish green 232. light purplish gray 32. grayish yellowish pink 99. strong greenish yellow 166. very dark bluish green 233. purplish gray 33. brownish pink 100. deep greenish yellow 167. vivid greenish blue 234. dark purplish gray 34. vivid reddish orange 101. light greenish yellow 168. brilliant greenish blue 235. purplish black 35. brilliant reddish orange 102. moderate greenish yellow 169. strong greenish blue 236. vivid reddish purple 36. deep reddish orange 103. dark greenish yellow 170. deep greenish blue 237. strong reddish purple 37. moderate reddish orange 104. pale greenish yellow 171. very light greenish blue 238. deep reddish purple 38, dark reddish orange 105. grayish greenish yellow 172. light greenish blue 239. very deep reddish purple 39. grayish reddish orange 106. light olive 173. moderate greenish blue 240. light reddish purple 40. strong reddish orange 107. moderate olive 174. dark greenish blue 241. moderate reddish purple 41. deep reddish orange 108. dark olive 175. very dark greenish blue 242. dark reddish purple 42. light reddish brown 109. light grayish olive 176. vivid blue 243. very dark reddish purple 43. moderate reddish brown 110. grayish olive 177. brilliant blue 244. pale reddish purple 44, dark reddish brown 111. dark grayish olive 178. strong blue 245. grayish reddish purple 45. light grayish reddish brown 112. light olive gray 179. deep blue 246. brilliant purplish pink 46. grayish reddish brown 113. olive gray 180. very 247. strong purplish pink 47. dark grayish reddish brown 1 14. olive black 181, light blue 248. deep purplish pink 48. vivid orange 115. vivid yellow-green 182. moderate blue 249. light purplish pink 49. brilliant orange 116. brilliant yellow-green 183. dark blue 250. moderate purplish pink 50. strong orange 117. strong yellow-green 184. very pale blue 251. dark purplish pink 51. deep orange 118. deep yellow-green 185. pale blue 252. pale purplish pink 52. light orange 119. light yellow-green 186. grayish blue 253. grayish purplish pink 53. moderate orange 120. moderate yellow-green 187. dark grayish blue 254. vivid purplish red 54. brownish orange 121. pale yellow-green 188. blackish blue 255. strong purplish red 55. strong brown 122 grayish yellow-green 189 bluish white 256. deep purplish red 56. deep brown 123. strong olive green 190. light bluish gray 257. very deep purplish red 57. light brown 124. deep olive green 191. bluish gray 258. moderate purplish red 58. moderate brown 125, moderate olive green 192. light bluish gray 259. dark purplish red 59. dark brown 126, dark olive green 193. bluish black 260. very dark purplish red 60. light grayish brown 127. grayish olive green 194. vivid purplish blue 261. light grayish purplish red 61. grayish brown 128. dark grayish olive green 195. brilliant purplish blue 262, grayish purplish red 62. dark grayish brown 129. vivid yellowish green 196. strong purplish blue 263. white 63. light brownish gray 130. brilliant yellowish green 197. deep purplish blue 264. light gray 64. brownish gray 131. strong yellowish green 198. very light purplish blue 265. medium gray 65. brownish black 132. deep yellowish green 199. light purplish blue 266. dark gray 66. vivid orange yellow 133. very deep yellowish green 200. moderate purplish blue 267. black 67. brilliant orange yellow 134. very light yellowish green 201. dark purplish blue

22 Understanding Color Names

general way, by the whole public" (National adverbs and adjectives. Figure 3-4 lists the Bureau of Standards n.d. a, 1). The ISCC-NBS hierarchy of qualifiers developed by Deane B. system is intended to be sufficient for the task Judd. of cataloging ten million colors, computed to The ISCC favored constructions such as be the number that the unaided human eye can grayish yellowish pink (color no. 32), intended differentiate. as a replacement for color names such as salmon The system identifies 267 major classes of or bisque. Which is superior is debatable. color, each of which has a name (figure 3-3). People usually recognize the subset of color to Numbers indicate distinctions within a class, a which salmon applies, having learned to necessary device because ten million names are recognize it from seeing the fresh, canned, or not available. In response to Gathercoal's smoked fish in restaurants, fish stores, rallying cry, the offending blackish white has supermarkets, and delicatessens. Grayish gone its way. Blackish red, reddish black, and yellowish pink is a less accessible commodity: other relatives escaped the purge. The project the ISCC color can be seen only by obtaining had no consulting epistemologist who might the appropriate color swatches. The question is have inquired what -ish contributes when less whether the public can understand the appended to a color name, as in reddish, ISCC-NBS system than how the system can be blackish, or pinkish. useful to either specialists or non-specialists in Larger in scale than other similar endeavors, its present form. the ISCC system illustrates the range of prob- An apparent intention is to replace color lems arbiters encounter in selecting appropri- names derived from object names (such as ate names for colors. Despite Gathercoal's call salmon) by compounds based on a minimal for clarity, the names are among the project's vocabulary. The ISCC stopped short of the lesser triumphs. As has been consistent among extreme, and the method reduces easily ad modern systems for simplifying color naming, absurdem. The system might consistently nonprimal names derived from object names eliminate, say, gray and pink because these were eliminated, with the exception of olive in colors are not elemental, although the names this case. Less typically, the ISCC abandoned are primal color names. Gray is a combination use of the more popular primals, such as red, of ; pink, of red and white. blue, or brown, as single names for major Grayish yellowish pink reduces to blackish classes of color. All subvarieties, except black whitish yellowish whitish red, which, in theory, and white, are identified by compound terms ought to mean the same thing. But color names in which the primals are supplemented by do not necessarily communicate more effect-

Figure 3-4. System of modifiers for ISCC color names. (After Judd 1979, 219.)

very pale very light very brilliant (very light, weak) (very light, strong)

pale light brilliant (light, weak) (light, strong)

weak moderate strong vivid (very strong)

dusky dark deep (dark, weak) (dark, strong)

very dusky very dark very deep (very dark, weak) (very dark, strong)

Understanding Color Names 23 tively if compounded from a limited repertory is the length of many ISCC color names. Light of common terms. Understanding moderate grayish purplish red is spelled with twenty- and blue does not explain moderate blue. three letters, three short of the length of the Carnap's theorem of the fundamental intel- alphabet. A decade before the ISCC project ligibility of color names falters when applied was begun, T. S. Eliot caused muttering about to the behemoth apparatus of the ISCC system. pedantry by opening a poem titled "The Hip- Not everyone imagines the intended color popotamus" with the elephantine poly- upon hearing grayish yellowish pink. The ISCC philoprogenitive. Even that infamously long prudently took this into account by recom- word contains 13 percent fewer letters than mending that a sample of the color be light grayish purplish red, probably intended exhibited, even to respondents familiar with as a replacement for color names such as gray, yellow, and pink. Exhibiting a large num- mauve or taupe. ber of samples is more to the point. If ten mil- By 1965 minor improvements to the ISCC- colors exist, each of the 267 ISCC classes, NBS system had provided what Kenneth L. including grayish yellowish pink, includes on Kelly proclaimed to be "the last missing link average thirty-five thousand discernible shades in our complete universal color language" of that color. (National Bureau of Standards n.d. a, 6). From Judd's system of modifiers for color names, the late nineteenth century onward, attempts incorporated into the ISCC system, relies on to systematize the naming of colors have typi- a hierarchical arrangement of ten words, any cally been introduced by the same announce- of which can be modified by very: pale, light, ment. The author or mover was galvanized brilliant, weak, moderate, strong, dusky, dark, upon confronting, in the morass of putatively and deep (see figure 3-4). The words are sim- unsatisfactory color names, one example that ple (everyone uses them), but not necessarily especially provoked ire (blackish white). This useful for identifying colors. Brilliant and vivid led to the revelation that the world, often the are synonyms or nearly synonyms, as are dark, no-nonsense manufacturing world, was in dusky, and deep, or pale, light, and weak. need of a more rational method for naming Moderate is poorly chosen in that the name colors. Never accomplished before, the task is moderate blue (color no. 182) implies that this now brought to fruition. The author is self- variety of blue is only moderately blue. The congratulatory, often with a lavishness more intention might reasonably be to identify the appropriate to discovery of an elixir for range as moderately light, moderately dark, immortality. moderately greenish, moderately purplish, or One of the utilities of the systems is to teach moderately grayish. Possibly it is moderately that the fabled difficulties in naming colors are something. It cannot, however, be moderately not so bad as we thought. To complain that blue, an inconsistency in terms unless further salmon is vague is injudicious if the best avail- explained. able substitute is grayish yellowish pink. Reliance on very, an unsatisfactory quail- Whether to apply science (or what is dignified fier for a color name, is similarly problematic, by that name) to the task of assigning names as in differentiation between pale violet (color to colors is an insufficiently aired question. I no. 214) and very pale violet (color no. 213). can muster no enthusiasm for a rational color Very is devoid of objective meaning. My opin- naming system if, without improvement in ion of what constitutes very (as in very nice) clarity, it purges language of the poetry of such can never exactly match yours. names as vermilion, viridian, cerulean, and tur- A side effect of the preference for qualifiers quoise. Clear or not, these mellifluous words 24 Understanding Color Names

are a pleasure to use, as grayish yellowish pink lowish pink improved on the despised topazy is not. They guide us in the right direction, yellow. toward the of color that inspires us to Putatively simplified color-naming systems observe it and learn about it. share an oddity in common with Esperanto. The portentous lengthiness of ISCC names None has been placed in wide usage, which is more than a sentimental issue. Color names, suggests the claimed simplicity is suspect.Even like all words, are used by including them in the developers of the ISCC system find it sentences. How is Homer's "rosy-fingered unwieldy for expository use, as can be seen dawn" to be recast by those who believe that in the voluminous writings of Deane B. Judd bureaucratic concatenations like very light pale (1900-72), president of the Munsell Color moderate reddish orange are of greater utility Foundation and associated for forty-three years than rose? A comparison of the ISCC system with the colorimetric section of the National with that of Albert H. Munsell, its predecessor, Bureau of Standards. Describing a scene the suggests color cataloging systems have grown reader is to imagine, Judd identified its colors progressively more ponderous in the name of as olive-drab, purple, green, blue, purplish- simplicity. black, blue-green, and pale green (Judd 1979, Munsell introduced his system in 1905, 485). Among these seven descriptive terms, proclaiming that "COLOR ANARCHY IS only pale green (color no. 149) and REPLACED BY SYSTEMATIC COLOR DESCRIP- unhyphenated purplish black (color no. 235) TION" (Munsell [1905] 1961, 24). Munsell are bona fide ISCC-NBS color names. found intolerable the continued countenanc- The lack of user-friendliness in the ISCC- ing of such notations as topazy yellow and NBS system, and to a lesser extent in earlier aimed to eliminate ambiguous color names. In systems by Munsell and Ostwald, derives in contrast to the 267 major colors of the Inter- part from the wanton discarding of widely Society Color Council,Munsell's list has 10. He known traditional color names. Either cerulean avoided qualifiers of the ilk of very and moder- or turquoise is a less ambiguous color name ate. No suffixes trail like cabooses: yellow than greenish blue. Vermilion means more green is preferred to yellowish green. than yellowish red or red orange. The whole- A literature primarily by members of the sale discarding of color names derived from Optical Society of America documents the per- object names (the largest class of popular mutations that led from the leanness of Mun- names for colors) raises the question of why sell's familiar system to its more weighty, less these names are viewed with distaste. widely known, successor.' In Germany, simi- The most likely answer is that color names lar revisions of the Ostwald system led to DIN derived from object names are contracted 6164 (May 1962), a devised by similes, and high school English teachers Manfred Richter and used for German Indus- deplore simile. Walt Whitman is said to have trial color standards. addressed notes to himself warning "avoid In the American system, the drove of appel- simile!" Metaphor is held to be preferable, evi- latives derived from Munsell's terse blue green dently because simile implies pointing, a ges- includes ISCC vivid bluish green, strong bluish ture to which aversion is widespread. Every green, dark bluish green, very dark bluish infant is instructed not to point, as if doing so green, moderate bluish green, and so forth. were either stupid or impolite. Munsell, given his preference for brevity, Color, however, is the special case in which might have wondered about them, or about pointing (ostensive definition) is integral to whether a color name like ISCC grayish yel- adequate communication. Showing Chap- Understanding Color Names 25 man's color swatches for rufous and fuscous dom. Although devalued as not real color is the correct way to make clear the difference names, words such as eggshell, avocado, and between the two browns. Not pointing to the turquoise are commonly used and continue to samples is impractical. The niceties of etiquette survive. run counter, in this instance, to the more con- The submerged imperative is rarely or sequential need for communication. never obeyed, in the sense that nobody goes Although we are taught to regard simile as to a jewelry store to examine coral necklaces an inferior construction for such purposes as if uncertain about what the color name coral writing poetry, it supports a special logic in means. What the hidden imperative naming colors. The collapsed simile in any acknowledges is less an immediate need to color name derived from an object name take action than the nature of color itself: of implies a hidden imperative that is, so to speak, what is definitive in regard to it. We all know to the point. Lemon as a color name means that ultimately the only way is to look. more than just a color like that of a lemon. It Anyone unable to discover by other means means the observer should look at a lemon if what the color coral looks like could acquire uncertain about the exact color. The name the information by regarding a piece of coral implies, in other words, the ostensive. as a color sample. The utility of the imperative Color names drawn from object names are in the simile is expressive, and it expresses finely tuned to their purpose, because each is what we know to be true. The bottom line for a reminder of where to look for a sample of color is the need to look, at some point, in the color. The nuance is sensed in folk wis- order to learn anything about it.

CHAPTER 4 The Limits of Language and the Logic of Color

We are up against trouble caused by our way of expression. Ludwig Wittgenstein, The Blue and Brown Books

anguage is used to report perceptual ex- not necessarily mean speakers will become perience, and all languages are arbitrary refined in observation of that phenomenon. or conventional. In what came to be Thousands of color names exist in the English called the Whorf-Sapir hypothesis, Benjamin language. More limited vocabularies for Whorf and Edward Sapir argued that perception identifying color are reported for some may be limited or controlled by the forms of non-IndoEuropean tongues. No doubt we have language (Sapir 1921, 1949; Whorf [1956] many more names for colors than the Eskimos 1967). A much-cited illustration tells us that the have for snow. This large repertory does not Eskimos have dozens of words used to inspire speakers of English to become describe types of snow. Because this must ena- especially expressive on the topic of color or ble them to find, so to speak, more to say, confident that they have much to say. Eskimos ought to be more sensitive to snow Sensibility to color among the peoples of the than others with more limited vocabularies. world, judging from the use of color in their art, I have never discussed snow with an has no correlation with the number of color Eskimo. Without prejudice to their knowledge names in their individual languages. of the topic, generalizing broadly along this To assess the effect of language, more has to line is unwise. That a language includes many be considered than gross numbers of words. words to describe a given phenomenon does We need to know how many speakers of the

26

The Limits of Language and the Logic of Color 27 language know the words, how often the tions of language when dealing with color vari- words are used, and how clear the meaning of ations too subtle to translate into words. each word is. The previous chapter contains Whether or not language directs per- a list of twenty-four common names for vari- ception, people use it to talk about color. eties of orange-red. Few people use more than Speakers of English can include color names two or three, because these words are ambig- in two classes of statements. The first refers to uous as color labels. No consensus exists, even color only nominally. Although the word color among experts, about exactly which shades of and individual color names may be uttered, orange-red ought to be called, say, vermilion. statements are being made about words or The lack of consensus reflects subjective ele- about rules of language. Consider the propo- ments built into all color names. sition that color is the name of a class contain- Vermilion originally meant a certain red ing particular colors, including black, green, pigment. Later, the name was applied to any red, and so forth. The assertion is not about color resembling that of the pigment. How color, but about how its name should be used. close a resemblance is required is anyone's It conveys to the listener that the rules for the guess. Pompeiian red, another name for a vari- English language are violated if the word color ety of orange-red, must have been coined by is used to identify the class containing all chairs somebody who visited the Villa of the Myster- rather than that containing all colors. ies at Pompeii.A large fresco in that villa shows Statements of the second type refer prop- human figures against a red background that erly to color as an object of perception. They has faded since the villa was excavated. Pom- pertain to the viewer's visual experiences. The peiian red means a red resembling-how criterion for distinguishing between statements closely?-the red of the fresco. For those who about language rules and statements about have never visited Pompeii or who never saw colors rests in the domain in which verifica- the red of the fresco before it faded, the color tion can be sought. If color is the name of the name means the red of the fresco as it appears class that includes red and blue among its in color reproductions in books, or the red of members and if these individual members are a swatch labeled Pompeiian red. said to be colors, confirmation can be found Speakers familiar with the etymology of the by observing how people talk. A person asked words might think of the pigment or the fresco to name colors is likely to mention green and when using vermilion or Pompeiian red. Many yellow, but not tables or refrigerators. speakers of English do not know the history of A statement about color as object is veri- these words. Vermilion means a color iden- fied by considering how colors look. The task tified to them by that name or any similar requires that the observer monitor his or her color. perceptions. To test the assertion that black is One effect of this vagueness of language is darker than white, I look at the colors to deter- that exact shades of color are difficult to iden- mine whether this is how they appear to me. tify in words. Does this affect our perception Because the domain of verification is visual, I of fine color differences? I think not. People pay little or no attention to what other people understand the futility of describing a color as, say. No statement can be verifiable in both say, a somewhat pale vermilion with a slightly domains at once, because no language rule is blue-gray tinge. When an exact match for a intended as a statement about objects or about color is important, color swatches or samples human perceptions of them. Nor are state- are used. People understand the problem and ments of the second type conventional, even the visual solution. They sidestep the limita- in instances where nearly everyone agrees. 28 The Limits of Language and the Logic of Color

That black refers to one color and white to We need not, and do not, go around ask- another or that dark has a meaning is conven- ing others whether they have a concept of tion. For persons of normal eyesight and com- color or demanding that the claim be substan- prehension to report that black looks darker tiated by those who say that they do. The per- than white is not, properly speaking, conven- son with a concept of color indicates, through tion, but rather, consensus. Convention speech and action, an awareness that color can changes when people develop new ideas about only be experienced visually. No individual how to talk or behave. For consensus to can be said to have a clearly defined concept change, people must acquire new beliefs about of color if wanting in this insight. A blind per- what they understand to be true. son might be said to have a concept of color, A theory that language forms affect the albeit a concept limited by the condition, if forms of perception amounts to a proposal that aware that color is a phenomenon the blind consensual opinions about perception (about, are unable to experience. Even that small wis- say, the perception of color) might be affected dom would save the individual from the error by a change in convention: a rose by any other of Aristotle's hypothetical blind person, who name might not be a rose-or it might be, if insisted on arguing about colors with those we consider the passing into obsolescence of who could see. individual conventions for color naming. The Relying on this criterion (which is a colors once called cochineal and aureolin look criterion for appropriate behavior), tests can no different today, although rarely identified be devised to determine whether a person by the names that were formerly used. behaves as if he or she has a concept of color. I can imagine a test in which the person is asked to name the color of some object he or Having a Concept of Color she has not seen, perhaps a table concealed in Language rules govern use of the word color another room. To pass the test, the subject and of the names of individual colors. Each must not try to answer before doing at least name refers to a circumscribed range. Red is one of three things: look at the object, receive no synonym for green. Light blue does not a report from another person who looked, or properly describe objects that can be called acquire access to a surrogate for first-hand vis- dark blue. Purple is not to be equated with ual experience, say, a color photograph. black. A complete listing constitutes a subset A color photograph would not help a blind of the catalog of rules for arranging words in observer answer the question, because sentences. cameras cannot perform the human function A catalog of this type can never explain of looking, any more than computers can per- everything people believe about the use of lan- form the human function of thinking. Cameras guage. Using words, including color terms, do, however, produce photographs at which correctly implies not only constructing proper human observers can look. Whether the per- sentences but uttering these sentences under son answers the question about the color of circumstances considered appropriate. It the table is irrelevant to the test. The critical implies the world external to language in point is that he or she show awareness that which these appropriate circumstances occur. some questions about color are requests for We routinely rely on this criterion when mak- reports about visual experience. They can be ing judgments about whether a given answered only by determining what the colors individual has a concept of color, a reliable of objects look like. understanding of its nature. The imaginary test is tempting to dismiss

The Limits of Language and the Logic of Color 29

as trivial or nonsensical. It offends our hope to be approximately similar for all. A prefer- that a concept, including a concept of color, ence or distaste for any individual criterion is more grand, more complex, than an insight might develop within a society, thus becom- on such a primitive level. Despite this objec- ing a convention of that society. But a sorting tion, the proposed test is consistent with the criterion is not a perception. manner in which people judge one another. Hans Hahn (1879-1934), a member of the Imagine a lecture on color by a speaker with Vienna Circle, commented extensively on outstanding credentials who was regarded as color while discussing language and logic. He unusually erudite. The listeners would not be attributed to convention the practice of call- impressed if they detected in this lecturer a ing some objects red and others blue ([1933] lack of awareness that questions about the 1966, 222-35). Hahn cannot have understood colors of objects cannot be answered until the the nature of the convention if he meant we objects are seen. We take it for granted that arbitrarily bestow the name blue as fancy dic- color is a visual phenomenon, that no concep- tates. Blue objects are called blue because they tual understanding of its nature can exist in display that characteristic; red objects are red absence of this acknowledgment. No person because they are that color. Conventional with normal powers of reasoning fails to color names are conventionally used to report understand that color is something we see. perceptually apprehended color differences. Color is also something at which, if not They are incorrectly used under other circum- prevented by visual impairment, we can stances. If asked to identify the color of a cof- choose to look or not look. Looking with fee cup, I look at the cup to make the insufficient care is commonplace, a lapse rarely determination. It would be unconventional to treated charitably if noticed. I lose confidence give "red" as an answer if was in theoretical proposals about color if the the- visibly blue. orist does not use his or her eyes or makes Because names are arbitrary and conven- statements that conflict with the evidence I tions change, blue objects might be called red, obtain by using my own eyes. Would anyone, and red objects might be called blue. A lan- to give the simple-minded example, trust any guage would not likely evolve in which the of the ideas about color presented by a per- name red was used for some red objects and son who insisted that white is darker than some blue ones and blue was used for some black? red objects and some blue ones. No one would know how to use such a language without information about which red objects and Sorting Criteria which blue ones belong in each class. Perhaps Visual experience, although not conventional all large objects could be called red and all in itself, is reported (and, prior to that, sorted) small objects blue, irrespective of redness or according to conventional formats. Bright red blueness. In that case, one criterion for sort- and bright green can be called similar if the ing (redness, blueness) would be replaced by sorting criterion is brightness, dissimilar if it another (largeness, smallness). The substitution is hue. According to other criteria, the viewer is similar to classifying people on Monday may legitimately conclude red and green are according to height, on Tuesday according to similar in some ways, dissimilar in others. yearly income. Red, in such a language, would By varying the sorting criterion, a different mean large, or a large red or blue object. form of interpretation (or a different empha- One of many color games proposed by the sis) is imposed on a visual experience assumed philosopher Ludwig Wittgenstein turns on an 30 The Limits of Language and the Logic of Color ornate articulation of sorting criteria: Life, as Wittgenstein argued, is gamelike. But human beings play multiple games simul- Imagine this game: A shows B different taneously. Knowing that red and green repre- patches of colours and asks him what sent patrician status has no bearing on the they have in common. B is to answer by separate awareness that the colors do not look pointing to a particular alike. Anyone taught to play Wittgenstein's sec- . . . . If in this game A showed B a light ond game would not be impeded from also blue and a dark blue and asked what they continuing to play the first. Everyday had in common, there would be no experience offers several parallels. Although doubt about the answer. If then he red and green are not indicators of patrician pointed to pure red and pure green, the social status, convention associates them with answer would be that these have noth- Christmas in societies with a Christian popu- ing in common. But I can easily imagine lation. A conviction that the two colors belong circumstances under which we should together as a symbol for a holiday season will say that they had something in common not persuade that they also belong together if and would not hesitate to say what it the criterion for sorting is hue similarity. was. Among sorting criteria appropriate to other Imagine a use of language (a culture) contexts, red and green traffic mean stop in which there was a common name for and go. A red light on an elevator indicates the green and red on one hand, for yellow elevator is going down. Adhering to whatever and blue on the other. Suppose, for interpretation is contextually appropriate has example, that there were two castes, one less to do with how the colors are seen than the patrician caste, wearing red and with the sorting criterion of the moment. Red green garments, the other, the plebian, can imply anger, virility, or communism. wearing blue and yellow garments. Both Green suggests envy, inexperience, money, or yellow and blue would always be Saint Patrick's Day. Green can point to the referred to as plebian colours, green and Irish or any group that adopts the color as its red as patrician colours. Asked what a emblem. It can suggest the environmental con- red patch and a green patch have in cerns of the German political party that calls common, a man of our tribe would not itself the Greens. hesitate to say they were both patrician Wittgenstein favored the view that (Wittgenstein 1958, 134). perception-not just interpretation--is con- ventionally determined, therefore easily sus- In the course of his game, Wittgenstein ceptible to modification. He argued that if, say, switched rules by changing the question to be light blue and dark blue were known by the answered. Initially, the player must determine dissimilar names Oxford and Cambridge, peo- what red and green look like. Later, the con- ple would say they saw no similarity between sideration is who wears these colors. In the them. The generalization is too broad. Names first instance, the report is about perceptions. can create a bias for or against the named In the second, the player supplies an interpre- object or color, which is why daffodil yellow tation by telling what the colors represent. sounds more appealing than pus yellow. But They represent, in Wittgenstein's game, either pairs of color names as different in sound and the modality of the patrician or the colors of spelling as Oxford and Cambridge exist. clothing worn by that caste. Viewers shown the pairs of colors, or familiar The Limits of Language and the Logic of Color 31 with them, recognize a visual similarity. Dark neighbor subscribes to terrestrial philosophy brown and chestnut, cerulean and turquoise, journals. The journals convey an impression that vermilion and crimson, ocher and mustard, are earthly colors all come in pairs, like the animals examples. in Noah's ark. Red finds its counterpart in Names for similar colors need not be similar not-red, blue in not-blue, and so forth. The words, a special case of the language epistemologist has not been able to locate an convention that synonyms and near synonyms example of not-red, a matter of concern. If need not be spelled or pronounced similarly. In many are available, an equal most cases they are not. The flaw in number of shades of not-red should exist. The Wittgenstein's argument about Oxford and epistemologist has begun to suspect the range of Cambridge blue is the assumption that a colors does not exist, but wonders, if this is the determination can be made about whether case, why it provokes lively discussions. She colors look similar by investigating whether wants to test the hypothesis that some they have similar names. This is no more discussions about color are not about color in reasonable than asserting that the paintings of any visual or experiential sense. They reflect a Monet must be similar to those of Manet or that semantic confusion in which questions about Austria should resemble Australia. The proper language rules are mistakenly classified as domain of verification is visual, because questions about visual experience. statements about how colors look are The philosophical literature on not-red is conventionally understood to be reports about more extensive than that on not-blue, notyellow, perception. They are not properly understood or not-taupe. It seeks criteria for distinguishing if thought to be statements of language rules. not-red from red, evidently of greater importance than distinguishing red or not-red from yellow. Little is in print about Red and Not-Red distinguishing red or not-red from not-chickens. Defining a concept of color as an awareness of Questions about the not-colors, several its visual nature amounts to categorizing color examined by Wittgenstein, usually concern the as a percept, to be encountered by experience. circumstances under which they can be seen: This implies that knowledge of color is a "Could it perhaps be imagined that where I see posteriori, after the fact. Nobody can know (or blue, this means that the object I see is not blue . nobody can know firsthand) about experience . . ?" (Wittgenstein 1967, 103). "But may not he or she has not had. The nature of the someone who is observing a surface be quite perceptual world cannot be anticipated, though preoccupied with the question whether it is we can and should think about what was seen going to turn green or not green?" (Wittgenstein after the act of perception. We cannot assume 1967, 19). that what we see will conform to what reason A viewer may indeed become pre- suggests. Human ability to reason, like human occupied with whether a surface "is going to language, has limits. One limit is that we are turn green or not green." But asking whether unable to see the future. Nobody knows the or not green is the color the surface will color of a color swatch sealed in an envelope become is syntactically preferable and less until someone sees it. likely to cause confusion. "The surface is not Consider the puzzlement of a hypothetical green" means the surface is an unspecified epistemologist who lives on an unnamed color, which only subordinately cannot be planet in an inaccessible galaxy. This distant specified to be green. As Bertrand Russell 32 The Limits of Language and the Logic of Color

pointed out, "When, as a judgment of Blue is not adequately described as not-red, perception, we disbelieve `This is red,' we are because blue is also not green, not black, not always perceiving that it is some other color" silver, not mahogany, and so forth through the (Russell 1948, 124). full range of colors. An explanation of the Not-red is a nonvisual concept, one of many difference between red and not-red cannot lead that make reasoning about color more difficult to greater understanding of what red means. It than it otherwise might be. Whether is a might produce insight into what the word not sensible term, in harmony with the nature of the means, but reveals nothing about color by so phenomenological world. Because I am unable doing. to anticipate the future, I might reason- If not-red were genuinely a recondite ably wonder whether the next car to come philosophical issue, it could safely be left to over the hill will be green. Interjecting fancies Wittgenstein and a few other specialists. I about not-green surfaces and not-green regard it as no philosophical question at all, just automobiles adds nothing of substance to the one of many examples of a looseness of question and sows confusion about its nature. expression in talking about colors that drifts into The not-colors apparently arose through the academic world from everyday life. The false analogy with the proposition, in logic, that looseness is traditional, a matter of any statement is either true or not true. The acculturation. We are not encouraged to logical proposition is often illustrated by discriminate carefully between statements about examples that use color names: Either it is true the perception of color and statements about that the table is red, or it is not true that the rules for using words. table is red. The coupled statements are not The bias is at a cost. Many intelligent, interchangeable with that other familiar pair: educated people, though nothing is wrong with either the table is red or the table is not red. The their eyes, never learn to think visually, to first set of statements concerns the truth of a reason in an orderly manner about visual proposition. The second concerns the color of phenomena. They become distracted by an object. semantic issues without recognizing that the In the first set of statements, "not true" has a issues are semantic. When not-red is accorded meaning, because not true is synonymous with the status of a color, as in, say, comparing it false. What "not red" means, other than with red, a parallelism is mistakenly assumed "unspecified," is less clear. The term implies between colors and logical classes. Logic, a classes as divergent as all colors in the world proper tool for determining the truth or falsity other than red or all entities in the world other of a statement, cannot be used to make than the color red. That the table is not red is a determinations about the colors of objects. It trivial truth about its color. It hides the lacks a facility for dealing with the visual necessary antecedent, the greater truth, that the continuum: with what we see or with what table is yellow, purple, light blue, or any shade objects look like. that we do not call red. Not-red might be an amusing piece of Considered as units in a classification nonsense were it not taken too seriously too system, each of the not-colors is an empty often. Hans Hahn found that "nothing is both category, devoid of attributes other than its red and not red. This is the law of contradic- emptiness and its name. One category is tion"([1933] 1966, 228). The law of contradic- notred, another, not-green, and so forth. Not- tion states that A and not-A cannot both be true red, which is not the name of a color,is also not at the same time. Red and not-red are poor self-sufficient as a label for a color condition. examples of its operation. Some objects are The Limits of Language and the Logic of Color 33 neither red nor not-red because they are partly opposites, the units in each pair linked in a red. A table painted with blue and red stripes is symmetrical complementarity that mimics the not properly called a red object: it is only partly complementarity of logical classes. red. Nor is it properly called not-red: it is only William M. Ivins, Jr., pointed out that, partly not-red. "intuitionally the Greeks were tactile-minded . . Color, which by nature is a continuum, is not . . Whenever they were given the choice structured in terms of parity. For several between a tactile or a visual way of thought, reasons, no viable classification system can they instinctively chose the tactile one" (Ivins begin with the premise that any object is either 1946, 9). The Greek conception of a primary red or not-red. A red/not-red system assumes substance, or matter, is even "the reduction of that all red objects share the same color, a the tactile-muscular intuitions to a sort of basic wrongly drawn parallel with the assumption, in philosophical principle." logic, that all true statements share the same The Greek conception of logic, which truth. More than one shade of red occurs in the remains our conception, is cut from the same experiential world. Language, more refined than pattern. If an intimation of truth, it resembles logic in this case, has forms that facilitate the truth of the hand. Exploring the world, the comparison among different reds. To say that hand discovers two states. It bumps into things "the red of the first necktie is very different or it does not, with no other possible condition. from that of the second, but almost matches that Matter is not space. In metaphorical extensions of the third" is syntactically allowable and of this given, what is true is not false, and right meaningful. is not wrong. Switches are on or off. Answers are yes or no. In the age of the computer, reasoning based Color and Logic on what Ivins calls tactile-muscular intuition What do they mean, these discussions of notred, nevertheless remains the state of the art. of Wittgenstein's surface that might turn Classical ideas about the emptiness of space and not-green? Like the imaginary epistemologist the non-emptiness of matter have gone their from another galaxy, none of us has ever seen way, both the ideas of the classical Greeks and not-red or not-green. At best, each term is those of classical (Newtonian) physics. But the meaningless, a negative definition. If color is assumption that the natural world exhibits something we see, inventing colors that cannot parity-that its building blocks come in be seen serves no useful purpose. I can go to a pairs-continues as a cornerstone of quantum hardware store and ask for the paint that is mechanics and other branches of modern not-red, though this is saying too little. I do physics. The assumption is rarely helpful at any better, maintain more , if I ask for the level in relation to color or . color I want. Consider the familiar assertion that black is If asking about the difference between red the opposite of white. Black/white is not a and not-red is nonsensical, why is it done? parallel construction to red/not-red. One set of Why does the question sound sober and famil- purported opposites pairs two colors. The other iar, as if serious issues are being raised? The consists of a single color contrasted with a answers lie, I think, in cultural attitudes about hypothetical entity. In each case, the asserted the visual sense that affect the way we reason oppositeness finds its paradigm in logic, which about color and about everything else. The holds that classes have opposites or comple- Greek philosophers passed along the bad habit ments. For every class K, there exists its com- of imagining the universe as a collection of plement, not-K. The logical assumption 34 The Limits of Language and the Logic of Color

supports the language rule that the word not aground at other points. Although we think of may be placed in front of any color name. The finite and infinite as opposites, mathematicians construction is semantically allowable. But is it have difficulty dealing with infinite sets. The meaningful? Does it lead to greater rules for handling them are not just the opposite understanding or just to nonsense? of rules for managing sets that are finite. Color As in the complementarity assumed between is obdurately resistant to analysis in terms of color and form, or attributed to colors opposite parity. Ideas about it that are incoherent, one another on the , the ubiquitous nonvisual, often begin with an assumption that assumption that black has an opposite (for every one aspect of color is opposite to some other. K, a not-K) has not produced consensus about Ivins traces to the Greeks our cultural where to find that opposite. For the philosopher preference for reasoning in terms of polarities, Daniel Sommer Robinson, "the opposite of for tactile concepts at the expense of those that black, is not, as we usually think, white, but are visual. But the devaluing of visual non-black . . . nonblack is the absolute or exact experience also has Judeo-Christian opposite of black" (Robinson 1947, 25). antecedents. The most influential aspect of Because black is often characterized as not a monotheism is not, I think, the idea of one god. color, its opposite can as reasonably be the This god, importantly, is invisible, suggesting doubly negative notnot-a-color. Beyond the that things we are unable to see can outweigh in merit of the several answers to the question, importance what is seen. Historically, the some more existentially absurd than others, the assumption contributed to the devaluing of difficult assumption is that colors can be visual sensibility-or affirmed a devaluing that presumed to have opposites, as if they were had already occurred. The Bible does not logical classes. provide careful descriptions of what things look The statement "black is the opposite of like. white" is a contradiction in terms. It says that All sense experience, not just visual the opposite of a given color is another color. If experience, eventually came to be devalued. black and white are not colors, it says that the The thread runs through Western philosophy, opposite of a not-color is another notcolor. which tells us the senses are unreliable. I sus- "Black is the opposite of not-black" creates the pect this means that the eye, if able to see cor- need for a definition of not-black. If not-black rectly, ought to be able to see God. In the West, (or nonblack) is an umbrella term covering we have not had the counterbalancing every color except black, we have said that the concept of the third eye. This inner eye sees opposite of a color is all other colors. invisible worlds, the metaphysics of what If not-black means every item in the world might lie beyond surfaces. It leaves to the other than black, the sense of the assertion is the external eye the task of understanding the truism that nothing is anything except itself. If world that is visible. not-black is a hypothetical construct (not a percept), then the opposite of a color is hypothetical. This last has the virtue of Doubting Colors suggesting that colors, like hummingbirds, may The deadening of visual sensibility displays not have opposites. itself in many ways. We rarely have a strong The tendency to structure thought in terms of sense of when words refer to colors and when sets of opposites underlies nearly all confused they refer to something else. A search under reasoning about color. But it also runs color in the card catalog of any library of The Limits of Language and the Logic of Color 35 respectable size turns up a literature that con- normal world in which objects show their siders the difference between the statements "it "real" colors. Imagine asking the observer to looks red" and "it is red." Nothing is called red decide once and for all whether the object that that never looks red, a rule of language. So "looks red" is red. To be certain, the observer anything that is red will look red to at least one would want to know the real color of the viewer on at least one occasion. Does the object, to see the object under normal or aver- difference between the two statements tell us age , and so forth. These and all similar about color or the perception of color? I think conditions cannot be met. They assume a world not. It reveals societal value judgments, again that does not exist, that has a stability lacking expressed primarily through rules of language. in our own. "It looks red" is a psychological report, "It looks red" is a linguistic mistake, though equivocal about the color of the object but def- traditional. It expresses a lack of harmony with inite about the observer's state of mind. The the existential world of visual experience. It observer is not certain about the color of the cannot help us to understand what we see, object. "It looks red" is the functional equiva- because it effectively asserts we ought to be lent of "it looks red, but I'm not sure," or "I seeing something else. The construction is too think it's red, though I might be mistaken." The deeply embedded in language to be easily viewer's uncertainty reduces to uncertainty eliminated. Its appearance in poems is instruc- about a logical proposition. The statement that tive. In Inferno, Dante described the Rock of the object is red is either true or false. Isn't the Purgatory, red in the rays of the setting sun. Is answer one or the other? How can anyone look the rock red, or does it just look red? and not know? T. S. Eliot's youthful improvements on In what has been called the tie-shop argu- Dame's passage in "The Death of Saint Nar- ment, Wilfred Sellars has theorized about why cissus" included the addition of a gray rock and uncertainty might arise. Objects look different firelight that reddens the red rock: colors under different kinds of light. A necktie that looks one color in a tie shop looks another Come in under the shadow of this gray color under daylight illumination. "It looks red" rock, implies that the object might not look red under And I will show you something different more optimal or normal lighting conditions. from either But optimal or normal conditions have not been Your shadow sprawling over the sand at established or do not exist. Because the color of daybreak, or any object varies with lighting conditions, no Your shadow leaping behind the fire object is the same shade of red under all against the red rock: conditions. Because illumination is never perfectly even, neither is any object exactly the How many gray rocks are in the Eliot passage? same color over all parts of its surface. An For a literary commentator, there might be two. object that looks red when the lights are on Eliot's red rock, as well as Dante's, might be looks gray or black in the darkness. If, in a regarded as a poetic compression. Yet I doubt visual or phenomenological sense, objects have we improve the poetry, or the sense, in either no absolute colors, what is the observer's case, by assuming the poet meant to say that dilemma about? the red rock just looks red. As a logical The issue is a societal expectation as old as proposition, if a rock looks red under certain the phrasing "it looks red." The construction conditions, it is red under those conditions. assumes the existence of an objective, real, or The conception of a rock that looks red but 36 The Limits of Language and the Logic of Color

really is not red is a semantic confusion possi- the conception, we need to assume a real or ble only in a society in which the evidence of normal world that reveals truths purportedly the senses is routinely mistrusted. As a visual hidden in ordinary phenomenological ex- phenomenon, color is of the moment, because perience. We do better, I think, to take the vis- we see (and live) only in the present moment. ual world as it is, freeing ourselves, where pos- Eliot's rock, as he said, is red when seen by sible, of ideas that refer to what it ought to be. firelight. It might be another color under other No other way exists of coming to terms with conditions. But those are other conditions. visual experience, including the experience of Looks like clouds the singular nature of seeing color. phenomenological reality. To make sense of

CHAPTER 5 Knowing How to Identify Color

Colour is something "out there," resistant to the eye. It is, so to speak, the vitality that shows upon the surface of an object, tinging as does the life-blood our skins. Adrian Stokes, Colour and Form

olor, among its several functions, can black, white, or silver with red, orange, and be defined as the name of a class. green, rather than with motorboats, cows, derby C This implies that the class can be iden- hats, or other items that are properly not colors. tified and the items belonging in it isolated. It In a refinement of the ability, few adults need not imply that everyone ought to know make errors in identifying black, white, brown, how to do so. Human beings, however, arrive or gray, as well as red, orange, yellow, and other at virtual unanimity regarding which items are major hues. The competence should not be colors. Because of the overwhelming taken for granted. People are not usually adroit consensus, a test for ability to separate the at estimating weight, velocity, temperature, or names of colors from those of noncolors would time of day. No person of normal eyesight and be fatuous (figure 5-1). Few test takers will comprehension fails to learn to identify colors. achieve less than perfect scores. The question of whether the ability can be lost is People rarely argue about whether red is a more clouded. color. Questions occasionally arise about black, white, gray, and the metallic colors. Even in these cases, I think most people would The Visual Agnosias make the reasonable choice. If asked to sort In a group of brain diseases called the visual colors from noncolors, they would group agnosias, perception and ability to speak are

37

38 Knowing How to Identify Color

Figure 5-1. Separating colors and noncolors. Object names that also identify colors are easily separated from those that do not. apricot brown charcoal cup grenade plate table avocado cabinet chartreuse curtain horse red vermilion banana camel chicken dog magenta rocket violet bed car child flag monkey ruler white black carousel church giraffe motorcycle saucer window blue computer gray orange shell yacht bomb chair cow green pitcher stool yellow

not generally impaired, but a condition arises tral items. Under environmental conditions, that is usually described as loss of ability to colors are often chosen (or sorted) for complex, understand the meaning of color names. indirect reasons. A customer in a cafeteria Agnosiacs follow a private logic of their own, selecting chocolate ice cream instead of , and their color-naming practices have been or tomato juice rather than orange juice, gives described by Norman Geschwind (Geschwind attention to color as an indicator of taste, the 1969, 4:98-136). The selective amnesias of the prime criterion in this case. patients assume different forms. One man, Agnosiacs apparently are tested for observed by Efron, substituted color names for understanding of only a few color names, object names, which the examiner believed had notably red, green, and blue, the colors of poker been forgotten. The man used silver as a name chips. This leaves unanswered the question of for keys, an incorrect reference to their material how the patients handled, if at all, words such or a correct reference to their color (Efron 1969, as turquoise, brick, salmon, scarlet, or lime. 4:137-73). Unless the term is defined with great care, what The larger group of agnosiacs of more type of knowledge (or behavioral skill) would immediate interest retains the ability to sort by be lost by an individual afflicted by amnesia for color, but conditionally. They cannot separate color names is unclear. Many or most words blue objects from yellow ones unless the task is used to name colors have more than one presented without using color names meaning. (Geschwind 1969, 121). Agnosiacs in this group By their nature, color names derived from are said to have no comprehension of what blue, object names function as color names only yellow, and other color names mean. Because when colors are what they are being used to we legitimately ask how a person would show name: lemon can identify either object or color, understanding of the meaning of color names, for example. Among the primal color names, case reports raise as many questions about the nearly all have additional meanings that are tests administered as about the patient's metaphorical. We speak of feeling blue, of performances. being in the pink, of seeing red, of being in the No standard psychological test has been red, in the black, or green with envy. We developed to evaluate understanding of color or understand the meaning of yellow coward, of color names, and is not easy purple prose, blue laws, green recruits, the to identify either. To bridge the gap in the social scarlet letter, scarlet women, or once in a blue sciences, ad hoc testing methods are widely moon. used that reflect the assumptions of the designer A patient who selectively forgets the mean- of the test. Little can be learned about ing of color names might show this in any of anyone's conception of color or understand- a variety of ways, each implying a different ing of color names by having the individual aberration. A person who uses the word olive sort colored poker chips or other similarly neu- correctly, except as the name of a color (or Knowing How to Identify Color 39

except as the name of an object), has a names. He never gave nonsensical answers selective amnesia for some meanings of some such as "the color of the poker chip is music." words. An individual who forgets all meanings The behavior suggests that what the patient of the word olive, because one of those remembered about color exceeded what had meanings names a color, may have an amnesia been forgotten. for all meanings of some words. If the Relying on data Geschwind provides, the condition is accurately described, the patient understood what a color is and had an individual, in an improbable pathology, would unimpaired concept of color. He knew that correctly understand the word apple (which color is a visual phenomenon and that objects cannot be the name of a color), but have no had to be looked at before answering ques- comprehension of orange (which can be). tions about their colors. Because even An agnosiac, in another possibility, may agnosiacs are not reported to be willing to sort have no difficulty using words such as olive, red poker chips from green in the dark, the whether for color name or object name. But, as patient was aware that determining the colors the literature reports, he or she may lack of objects requires the presence of illumi- understanding of blue, green, or yellow, again nation. an improbable pathology. Of immediate The patient also realized that colors have concern, the defect would be limited to names. He knew some of those names and forgetting the names of the primal colors. It apparently had no difficulty pronouncing them. cannot be regarded as an equivalent for He understood that color names were to be forgetting the names of all colors. chosen, from those he knew, in response to the The literature provides no way of examiner's questions. Importantly, he did not surmounting the absence of some kinds of data. use the same name for every color. He evidently Can the agnosiac remember what understood that one purpose in using different means when it refers to the weight of an object, names is to acknowledge differences among but not if a reference to the object's color? What colors. response would be elicited by instructions to The patient's difficulty was limited to an sort items according to whether their color was inability to correctly judge which color name, shiny or dull? If an agnosiac confused red with from among those he knew, would have been blue, would he or she describe sadness as most correct or appropriate. Geschwind's patient feeling red rather than feeling blue? Can an may have forgotten that color names are not agnosiac, or anyone else, remember how to applied to colors at random (blue is not called order a glass of Burgundy wine in a restaurant yellow). He may have lost the ability to sort or but forget that the color name burgundy refers to distinguish between an order that is random to the color of the wine? and another that is hierarchical according to Reports about agnosiacs, despite some criterion. conclusions to the contrary by the authors, He may have forgotten the sorting criteria rarely suggest the patients have forgotten color for determining which name went with which names. They appear to have forgotten, instead, color, as opposed to sorting criteria in general. the rules for determining which name to use. A Or he may have garbled the sequence, for patient observed by Geschwind gave wrong unknown reasons. The distinctions between answers when asked the colors of objects, these several failures in comprehension are substituting "red" when he should have said subtle. But they point to differences as vast as "green."' But the patient evidently those between not remembering a person's responded to questions about color with color name and not realizing people have names. 40 Knowing How to Identify Color

Because Geschwind's patient was able to ized areas of the brain that can be identified, judge, visually, the difference between red and color information poses a problem. How green, some disturbance in his sense of order or would the conception of, say, coolness be clas- appropriateness (choosing the right name for the sified or stored? Coolness refers to a condition circumstances) is implied. This is a selective of temperature, but we also speak of cool inability to use language correctly. Why color colors. names alone are confabulated is not clear, nor In English, though not in all languages, can we be certain this is the case. The literature words occur that are not considered names of does not suggest extensive testing to determine colors, though these words function whether agnosiacs also confuse names for syntactically as if they were. Colors can be animals, people, streets, or other items. A sorted according to whether, say, light, dark, or question that bears on societal expectations is . Yet light, dark, and pastel are not why it is considered abnormal to mix up the considered to be color names. Other terms names when asked to identify colors but not applicable to colors are subjective. A subject when asked to identify state capitals. can be asked to sort colors according to Geschwind, explaining the agnosias, whether, say, beautiful or interesting, which suggested "the patient with a colour-naming again are not names of colors. How are these disturbance can give only a poor account of his terms understood, and these sorting tasks colour-experience or indeed none at all since performed, by agnosiacs? If they cannot sort his speech area has little or no access to colors according to whether, say, light or information about the colour-experience of the beautiful, should the impediment be classified " (Geschwind 1969, 133). More as a forgetting of color names? Is a broader recent work with patients in whom right and difficulty in using speech implied? These left brain hemispheres have been separated by questions are for the medical community. Yet severing the corpus callosum (usually to relieve answering them requires a clearer understanding epileptic seizures) indicates that speech is of the nature of color names, how colors are localized in the left brain. Visual and spatial described in English, and how color-sorting understanding, required for understanding of tasks are performed. color, is localized in the right. Either Assume, say, that color information is hemisphere is capable of independent stored in two areas of the brain, one handling functioning, although not in the same way. visual information about how colors look, the Split-brain patients may be able to copy simple other, information about how words are used in geometrical forms with the left hand making statements about color. The difficulty (controlled by the right brain) more skillfully of agnosiacs seems to lie in the second area. than with the right. The area, however, spans out to include any Damage to localized areas of the brain has conception that might be used in sorting colors. been correlated with specific behavioral A subject can be asked to sort colors according malfunctionings. Injury to Broca's area, in the to whether, say, the color has been seen lower rear part of the left brain's frontal lobe, recently, is fashionable this year in clothing, or causes aphasia, difficulty in speaking or is used in traffic signs. inability to speak. Injury to Wernicke's area, Any word in the English language might be straddling the parietal and temporal lobes in the used in talking about colors or in giving rear center part of the left brain, leaves speech instructions about how they should be sorted syntactically normal but senseless. If, however, and labeled. The primal color names, although certain kinds of information are stored in local- they are the words most often used for label

The Munsell Color Tree. (Courtesy, Munsell Color, 2441 N. Calvert St., Baltimore, MD 21218.) The Munsell parameters for ranking color are hue, value, and chroma. The Ostwald parameters are hue, black content, and white content. In both systems, dullness or shininess of surface is acknowledged as a fourth quality, which cannot be incorporated into a threedimensional model.

The Munsell Value Chroma Chart. (Courtesy, Munsell Color, 2441 N. Calvert St., Baltimore, MD 21218.) The hue circle, although based on the solar spectrum, is an idealized and modified model of it. The spectrun. shows no red-violet, a color included in the hue circle. And the bands of color in the spectrum are not of equal width, though the hue circle shows color in equal sectors. White has traditionally been equated with light, and black with its absence. But the gray scale is as stylized as the hue circle. White light is brighter than any white paint, and the experience of seeing darkness is different from that of seeing black paint.

Combined Hue-Value Scale (after Kandinsky). (Illustration by Pamela Dohner.) Although the achromatic colors (black, white, gray) have no hue, the hues can be ranked according to value, their degree of lightness or darkness. In Kandinsky's arrangement, yellow falls near white because both are light. Blue falls near black because both are dark. Red occupies an intermediary position, and is neither very light nor very dark.

Combined Hue-Value Circle. (Illustration by Pamela Dohner.) In this combined hue-value circle designed by the author, black, white, and the spectral hues are ranked by degree of darkness around the perimeter of the circle. Each color is opposite a color of high contrast to itself, although no color is opposite to its complementary color. The strongest value contrast is between black and white; the strongest hue contrast, red and green.

Piet Mondrian, Composition in Red, Blue, and Yellow, 1937-42. Oil on canvas, 233/4 x 211/8 in. (Collection, The Museum of Modern Art, The Sidney and Harriet Janis Collection.) Over a period of years, Mondrian gradually reduced his palette to the five colors he regarded as pure: red, blue, yellow, black, and white.

Kenneth Campbell, Untitled, 1949. Casein on paper, 31.5 x 22.3 in (80.01 x 56.64 cm). (Photograph by Bob Rubrick.) Forms affect colors by limiting their areas. This hard-edge geometrical abstraction uses a relatively large number of relatively small shapes. The effect would have been different with, say, the same colors but only four or five large shapes. And it also would have been a vastly different painting if done entirely in, say, . The perennial question of whether form is more important than color has no answer, other than that a change in either is a significant change.

Betty Vera, Hot Planet, 1988. Cotton tapestry with painted warp, discontinuous weft inlays, and broken twill weave, 36 x 36 in (91.44 x 91.44 em). (Photograph by Betty Vera.) In this tapestry, the artist selected saturated middle-value hues that have little lightdark contrast. If reproduced in black and white, the design would almost disappear. The contrasts are of warm-cool, and of hue changes.

A1 Alcopley, Approach, 1987. Oil on canvas, 50 x 68 in (127 x 172.72 cm). (Photograph by Bob Rubrick.) A painting, like anything else we see, can be thought of as a matrix or aggregate of color spots, an arrangement of pigments on a surface. The color itself can become an image, icon, or subject.

Herman Cherry, Day and Night, 1987. Oil on canvas, 66 x 60 in (167.64 x 152.40 cm). (Courtesy of White Pine Gallery, New York City. Photograph by R. Cherry.) Like style, artists' use of color is individualistic. In this painting, day and night are associated not with black and white but with an arrangement in which pink takes the place of green, and all spectral hues other than green are included.

Nina Tryggvadottir, Untitled, 1956. Oil on masonite, 49.2 x 24.2 in (124.97 x 61.47 cm). Shapes are always referential, even those shapes that only refer to the forms of geometry. The irregular shapes in this painting are biomorphic; the austere blue-green, black, and white of this abstraction suggest a rocky landscape.

Jeanne Bultman, Istanbul-Nigbtfall Rising, 1979. Stained glass fabricated from collage by Fritz Bultman, 84 x 48 in (213.36 x 121.92 cm). (Photograph courtesy of Jeanne Bultman.) Media dictate their own conventions, limiting the ways in which colors and shapes can be used. Stained-glass color is filtered light, rather than light reflected from surfaces. To view it, light is required behind the work of art but not necessarily in front of it. The black lines between the colors are a necessity rather than an aesthetic choice; the channels of lead support the pieces of glass. Closed forms are the natural forms of stained glass, and of the paper collage on which this piece is based.

Mughal, Harivamsa (The Genealogy of Hari): Krishna Lifts Mount Govardhan, c. 1590. Opaque watercolor on paper, 11.375 x 7.875 in (28.9 x 20 cm). (The Metropolitan Museum of Art, Purchase, Edward G. Moore, Jr. Gift, 1928. (29.63.1).) The color combinations in this painting are related to those in El Greco's View of Toledo, but the moods are quite different, partly because of the difference in subject matter, partly because the color

Egyptian, Pectoral of Senwosret IL (Detail). Gold, amethyst, turquoise, feldspar, carnelian, lapiz lazuli, and garnet. Pectoral, 3.25 x 1.75 in (8.2 x 4.5 cm). (The Metropolitan Museum of Art, Rogers Rund and Henry Walters Gift, 1916. (16.1.3).) The colors in this piece of jewelry are the colors of metals and precious and semiprecious stones. Here opacity, translucence, and transparency come into play.

El Greco (Domenicos Theatocopoulos), View of Toledo. Oil on canvas, 47.75 x 42.75 in (121.3 x 108.6 cm). (The Metropolitan Museum of Art, Bequest of Mrs. H. O. Havemeyer, 1929. The H. O. Havemeyer Collection. (29.100.6).) Colors and deep shadows create a stormy, moody effect, perhaps because we associate the colors with natural phenomena and link these phenomena to emotions or states of mind. Subject matter can strengthen or dilute the effect. And some paintings more than others invite us to consider the state of mind of the artist who chose the colors and arranged them.

Renoir, Madame Charpentier and Her Children, 1878. Oil on canvas, 60.5 x 7.125 in (153.7 x 190.2 cm). (The Metropolitan Museum of Art, Wolfe Fund, 1907. Catherine Lorillard Wolfe Collection. (07.122).) Blue is said to recede, a characteristic of cool colors. But in this painting, the pale blue of the children's dresses is set off by the warmer tones of the background, and by its proximity to the strong black-white contrast of the mother's clothing and the dog.

Knowing How to Identify Color 41 ing colors, are never the only correct identifiers. mation about color experience implies, an The color labeled yellow can also be correctly analogue for the behavior of Geschwind's classified as light, bright, warm, or the color of patient is that of a normal person wearing a daffodils. A certain yellow may be lemon blindfold, compelled to make guesses at the yellow, chrome yellow, or any of the many colors of objects because the objects cannot be nonprimal names for different . The seen. Another analogue is a person asked to picture of the abilities of agnosiac patients label colors according to whether they were, should be filled out by determining their say, , taupe, or aureolin. If the person did general level of verbal competence, a neces- not know the meaning of these color names, sary background to assessing impediments in random guesses and errors might be made. labeling colors. In one conceivable type of Geschwind's patient apparently looked at impediment, an agnosiac might misuse, say, the the objects before answering, a significant color name yellow but be able to identify detail because it ties his behavior to the yellow as a light or bright color. The agnosiac behavior of anyone else. The patient is not might be able to describe colors by simile, reported to have said he could not identify the identifying, say, which color looked like the sky colors of the objects he had looked at. Nor did or which was the color of grass. he say he had forgotten color names. He lacked So far as is known, agnosiac patients have no awareness of having forgotten. Or he resisted physical reason for the specialized lack of accepting this as his condition, if it was his con- access Geschwind attributed to them. What is dition. needed is a clearer picture of what they lack Perhaps, like the rest of us, Geschwind's access to. The terms color experience or patient found it inconceivable that anyone could account of color experience are not easy to look at an object and be unable to name its color define. Is it an account of a color experience if I correctly. Chimpanzees who learned to use sign say my favorite color is blue? Because of the language, under the tutelage of Beatrice and structural and syntactical peculiarities of the Robert Gardner at the University of Nevada, English language, a selective amnesia for color were taught the signs for the principal colors names cannot occur. Therefore, it cannot be the and used them correctly. If the brain has a basis of a pathology, or it cannot be an accurate special method for storing words and concepts behavioral description of that pathology. If that cannot be forgotten or are almost never color names were to disappear from the English forgotten, the names of the primal colors, language, we would lose at the same time many though not other color names, might fall in this names of objects, many metaphorical group. Color might be an unforgettable concept. expressions, the de facto color names (pastel, It had not been forgotten by Geschwind's dark), and many conceptions applicable to color patient, which is why he was able to answer the but also used to label other items (beautiful). questions put to him. Although the criteria for unforgettable con- cepts have not been defined, we often assume they exist. Freud, writing on pain and pleas- Forgetting Color Names ure, never considered the possibility that any- The term forgetting is used arbitrarily, a one could forget the nature of these conditions looseness that bears on studies of memory. An or even what the words meant. Pain and agnosiac is said to have forgotten color names, pleasure precede language. Organisms low on but a patient with anorexia is not said to have the evolutionary scale move toward some things forgotten how to eat. As lack of access to infor- and away from others, actions we interpret in 42 Knowing How to Identify Color

terms of what the organism likes or dislikes. ness. The responses of Geschwind's patient are Mammals, say cats, react to pain and pleasure consistent with this. He knew what color was, with sounds and body language that human knew questions were being asked about it, and beings understand perfectly. Nobody mistakes a understood he was expected to answer the suffering cat or an angry cat for a contented cat. questions. He gave wrong answers. Pain and pleasure, we might theorize, are unforgettable because reaction to these conditions is integral to the life process. Memory and Value Judgment Organisms do not forget they are alive. Selective amnesia (or selective forgetting) is Describing pain or pleasure in words is an accepted as normal when it occurs in the general afterthought, a late development in the population. Everyday examples include evolutionary process. But color, or the light forgetting how to use logarithms or condition we call color, falls into the same remembering people's faces but not their names. category. Many organisms low on the The lapse implies a selective failure of will: an evolutionary scale move toward light or away aversion to paying attention to what does not from it, or selectively respond to light of certain seem important. colors. An elderly person with severe memory Whether plants can feel pain or pleasure is problems may forget what everyone considers cc a highly controversial question, often dis- important. A living relative may be addressed missed as ridiculous although some people talk by the name of a person known many years ago, to their plants. That plants respond to color and as if the conception of time had collapsed. A light, however, is firmly established and many person with Alzheimer's disease may forget of the mechanisms are understood. In, for what happened five minutes ago or may for- example, the process called phototropism, get the need to put clothes on before going out plants grow toward the light. Cells on the side in the street. Value judgment by an observer of the plant stem away from the light multiply is a central issue in memory, because strict more rapidly, the mechanism by which the societal rules determine what is allowably ju~ process operates. forgotten and what we expect will be That a plant might forget how to grow remembered. A job interviewer in a business toward the light is inconceivable. We would office may forget the names of unsuccessful job classify such a plant as a damaged plant or a applicants. That interviewer should not forget dead plant, not a plant with amnesia. Living the name of his or her employer. Nor should the organisms cannot forget how to respond to light, interviewer forget the name of his or her because they cannot forget how to live and still spouse. continue living. Because light is a different The absentminded, like the poor, are always color from darkness, a mechanism for among us. Although small boys have no discriminating between colors (or between light difficulty remembering the batting average of and darkness) is implied, even for tulips, major league ball players, an adult population amoebas, and other organisms that do not have perennially unable to remember how to use the eyes. The eye, like the brain, is a centralized names vermilion, carnelian, magenta, organ of great refinement that evolved to chartreuse, and mauve is not difficult to col- control decentralized mechanisms that preceded lect. Many people misuse, confuse, or forget it. color names other than the primals. The inep- I conclude we cannot forget what color is, or titude escapes notice because commonplace, forget, say, how to notice a light in the dark- an indicator of societal priorities. Few except Knowing How to Identify Color 43 students in art schools are actively encouraged word red to be used correctly by the patients, to develop a refined ability to name, observe but doubtless would have been more forgiving carefully, match, or mix colors. about, say, the words carnelian or cerise. Why? Those who misuse many of the less famil- Rules exist, and we follow the rules. But none iar color names shed light on what otherwise of us can say what they are or why we make presents itself as an uncanny aspect of the value judgments as we do. agnosias: unimpaired skill at sorting, combined What goes on in the minds of agnosiacs with inability to name correctly. Nobody may be no more curious than what goes on in needs to know the words vermilion and our own minds. I am less interested in why magenta to sort vermilion and magenta color women like color, if in fact they like color, than swatches. Naming and sorting are unrelated in the train of metaphorical association that skills, and we should not expect otherwise in suggests the interest would be appropriate. We agnosiacs. cling stubbornly to the bad habit of attributing Geschwind's patient who confused red and gender to what is neither masculine nor green displayed a different magnitude of feminine. The nearest analogue may be the impairment from that of an individual who alchemists, who superstitiously imagined that confuses vermilion and viridian. Whether or the union of chemical elements could be not red, like color, is an unforgettable concept, meaningfully compared to human coitus. The errors are not expected in use of the primal association never led the alchemists to any color names, which even signing chimpanzees insight of scientific significance. It catered to a learn to handle correctly. If encountered in the vainglorious idea, shared by both men and course of everyday affairs, an individual whose women: that the world is a mirror of human vision was not anomalous but who misapplied concerns. The cosmos, viewed in this manner, the words red and green would be regarded as reduces to a collection of opposites reflecting retarded. Yet gender-based stereotypes can- the oppositeness attributed to male and female. not be ignored and reflect societal value judgments. The man unable to remember unusual color Color Names and Object Names names is as familiar a figure as the woman un- Although thousands of color names exist, the able to remember the names of the parts of an average adult uses very few. Anyone asked to automobile engine. Interest in color passes list all the names of objects he or she could widely as either feminine or effeminate, a code think of and then all the names of colors would that questions whether it has a value, whether produce a longer list of object names than of remembering information about color is color names. Yet I doubt that seeing a differ- important. When a hostile art critic describes a ence between two shades of cerulean blue is painter as merely a , a work of art as more difficult than seeing a difference between merely colorful, a world of attitudes is con- a chair and a sofa. The disparity in vocabulary veyed by the word merely. This world of atti- mirrors human value systems rather than the tudes bears on the central issue in the agnosias. nature of the phenomenological world. Form With due respect for the strange symptoms of differences are regarded as more significant the patients, interpretations of these symptoms than color differences. are also strange. The patients are interviewed The bias, unexamined, infects scientific the- by medical personnel who share the expecta- ory, much like the outlaw computer programs, tions and misapprehensions about color com- called viruses, that infect and disrupt modern mon through our society. They expected the computer programs. Hermann Rorschach, 44 Knowing How to Identify Color

without examining his reasoning or its ing that the blueness of a chair is not the chair traditional sources, went so far as to suspect that requires prior understanding of what a chair is. only defectives and artists paid more attention An individual cannot correctly name colors but to color than to form when given the inkblot lack understanding of what an object is. test. Rorschach inkblot test scores are not Learning to identify colors, like learning to available for, say, Newton, Goethe, Munsell, or count, requires an understanding that objects Ostwald. Would they, too, have paid more can be considered to have attributes. I hazard a attention to color, at the expense of form, than guess that among those of low intelligence, low Rorschach regarded as appropriate? literacy, or low educational level, difficulty in As Rorschach realized, paying attention to naming colors and in counting exceeds color rather than form is unconventional, a difficulty in naming objects. deviation from societal rules. But the What is forgotten by patients with visual development of human thought depends on agnosias is less interesting than what is entertaining ideas that are different and new, remembered. Even agnosiacs know what a color that are not just a recyling of ideas from the is, which is why Geschwind's patient gave color past, not just a rehash of what everyone thinks. names, though the wrong color names, in The presumed propriety of preferring form to response to the questions put to him. Asking color is conventional, an old way that may not how agnosiacs and others make the be a better way or may not be suitable for all determination amounts to seeking an circumstances. A plant would be in trouble if it operational understanding of how color is decided not to grow toward the sun or if it identified as color. Three answers to the decided to grow in the direction of forms that, question are old enough to have been thought say, included right angles. upon by Plato and Aristotle. First, color is a The preference for placing form above visual percept. Second, color is either an color finds serendipitous support in constraints attribute of light or a phenomenon caused by set by language. Children learn the names of light waves. The third answer, related to the forms or objects before learning color names, first, is that color, as visual percept, is a route that cannot be avoided. Understand something other than visually perceived form.

PART TWO

Color and Light

We can take up the subject of the study of color in two quite different ways; namely, as a subject in physics or as a subject in art. If we consider it from the point of view of physics we are dealing with exact laws of light and optics, and may proceed on a basis of fairly well established facts. Walter Sargent, The Enjoyment and Use of Color

Whereas the world of naive men is somewhat confused, and reveals its subjective character in any critical discussion of its properties, in the world of the physicist no confusion and no contradiction are tolerated. Wolfgang Kohler, Gestalt Psychology

CHAPTER 6 Light of Day, Dark of N fight

There is a degree of light which surfeits, a want of it which starves the visual organ. Horatio Greenough, Form and Function Suppose, Ananda, there is no light and they are unable to see things, does that mean that they cannot see the darkness? If it is possible to see darkness when it is too dark to see things, it simply means there is no light; it does not mean they cannot see.

The Surangama Sutra

olor is caused by light, a familiar tioning of the color sense is more or less assertion. We explain what it means in dependent on the illumination. The brighter the either of two ways, according to light, the more color there is in the world" (Fucks (1908] 1924, 243). For Claude Bragdon, whether light is ordinary daylight or light as it is the underlying basis of theory was that "the sun described in the physical sciences. Neither is the affector, the eye is the receptor" (Bragdon explanation suffices if light is considered in the 1932, 122). For Robert Boyle, "the Beams of alternative sense. Neither addresses the curi- Light, Modify'd by the Bodies whence they are ous question of whether light exists in isola- sent (Reflected or Refracted) to the Eye, tion from color or apart from its own colors. produce there that kind of Sensation, Men commonly call Colour" (Boyle [ 1664] 1964, Light of Day 90). If light means daylight, objects exhibit many It is more difficult to agree that "if the colors during the daytime. They appear illumination sinks below a certain limit the indistinguishable from one another in color perception of colors fails altogether" (Fucks (although not colorless) at night. All cats, we [1908] 1924, 90). , inseparable are proverbially told, are black in the dark. For from vision, achieves widest scope (the greater the ophthalmologist Ernest Fucks, "the func- number of colors can be seen) within a circum

47 48 Light of Day, Dark of Night

scribed range. With sufficient light, we see Fuch's equating of brighter light with more many colors. Dimming the light reduces the color relies on the assumption that the number of colors perceived, though never to illuminating light is white or nearly white. The none. When light is absent, the color of the full range of yellows cannot be seen under a darkness remains. All looks black. light, however bright, that is blue. Beyond this, We do not know whether long-term bright light causes harsh shadows, and diffused deprivation of light causes permanent light is kinder to colors. Most color films for impairment of vision. Goethe argued that acuity cameras, like most human beings, distinguish improves, and that prisoners confined in the fine nuances of colors less effectively on a dark learn to see without light, at least dimly. sunny day at noon. A gray day provides bet- Excessive illumination is obviously harmful, ter conditions, as does morning or evening the reason for not looking into lights of blinding light. brightness or at eclipses of the sun. Color vision is also adversely affected by oscillation from one condition to another, the reason eye strain is Dark of Night caused by strobe lights. And changes occur in if colors are brought out by daylight or require the of human vision as a result of its presence to some degree, light, in this sense, aging or certain diseases. causes color. The causative relationship is said In his essay "Nature" Ralph Waldo Emerson to have been proved by establishing it (1803-82) found "no object so foul that intense scientifically. Nature's puzzles rarely unravel light will not make it beautiful" (Emer- easily, and the nature of proof, including son [1836] 1969, 8). Leonardo da Vinci scientific proof, has inspired extensive debate. (1452-1519) was less convinced that strong Setting aside those issues, scientific theory light flatters colors. In one chapter of his trea- explains, or proposes to explain eventually, the tise on painting, he finds "colors situated in nature of human experience in the everyday a light space will show their natural beauty in world, including the perceptual experience of proportion to the brightness of that light" seeing color. The presumption is that human (Rigaud [1802] 1957, 145). Elsewhere, "pol- beings are able to give an accurate report of the ished and glossy surfaces show least of their experiences science will explain. A reason for genuine color... The parts which receive the resisting that assumption in the case of color is light do not show their natural color" (Rigaud that many conventionalized ways of reporting [1802] 1957, 137). Light will not enhance the the perceptual experience are nominal. colors of objects if reflected from their surfaces It is not true, for example, that objects have in a manner that produces excessive sheen, no color in the dark. They simply all look the glossiness, highlight, or glare. These effects, same color, the reason they cannot be each of which can be given an optical expla- distinguished from one another or from their nation, consist of a greater than optimal environment. That is immediately fatal to the amount of reflected light. The excess may exist argument that light causes color or is throughout the field (glare) or over a small por- identifiable as its sole cause. The color of the tion (highlight). In photography, where high- dark of the night is seen when illumination is lights and glare are rarely desirable or not present. Furthermore, the color common to acceptable, methods of controlling them the environment and all nonfiuorescent objects include photographing through polarizing in the dark is poorly described as black. In many filters. cases, the tonality is so much closer to char- Light of Day, Dark of Night 49

coal gray that the term visual gray has been teristics of the visual gray. Colors in daylight used (Stokes 1937, 22). In astronomy, the seem to be located "out there," attributes of the corresponding phenomenon is called airglow, a surfaces of objects rather than of human narrower word referring to the deviation from neurological processes. Because no horizon line absolute blackness observable in the night sky. crosses it (we assume that horizons are "out One conjecture attributes airglow, described there"), the visual gray appears closer: not so as a weak glow, to stray radiant energy from the close as to be inside the eye or within the sun. Possibly, interpreting the night sky as eyelids, yet no more than a few inches or feet aglow is a rationalization of visual experience. away. To Stokes, the visual gray gave an Defeating any expectation to the contrary, the impression of being soft, perhaps because, like color of the sky at night is not black. Even in the blue of the sky, not associated with the rural areas removed from bright city lights, it is hardness of objects. only nominally black, which implies no more The deviation from absolute black, than that human beings have lapsed into the sometimes conjectured to be caused by habit of calling it black. spontaneous neural discharge, may be a function The nonblackness of the nocturnal sky- of the absence of what M. E. Chevreul called scape has a counterpart in the nonblackness simultaneous contrast. Colors are brought out by of enclosed rooms. Indoors or out, black light, but also by juxtaposition to other colors nights are not black, nor do all cats look black that look different. Black looks most dark (or in the dark. All cats look gray in the dark, even most black) next to white or next to light or black cats. Light is not a sole cause for color bright colors. Color contrast is absent in the if we see the gray of the darkness and do not darkness, and what is caused by absence of light cease seeing in light's absence. The most is absence of color differentiation, not absence intense, the most nearly absolute, are of color. not caused by the absence of light. They are The familiar question of whether everyone caused by its presence. The color of black pat- sees colors in the same way assumes its own ent leather and of other extremely dark light in this context. If one person objects-colors that are blacker than the black saw the darkness as gray, another as green, of night - are seen in daylight but not in the another as yellow, the fundamental nature of the dark. experience would be unchanged: nobody can The darkness is often mottled or reticulated, see objects in the dark because they look the sometimes in two tones of gray. Traversing it in same color as their environment, whatever that stately procession are minuscule spots or color might be or might be called. flashes of white, orange, and pale blue. These markings, called phosphenes by mod- ern researchers, were identified by Helmholtz Undifferentiated Color as the "self-light or intrinsic light of the retina" Fear of the dark is explained in terms of (Helmholtz [1856-66] 1962, 2:12). They childish imaginings of demons and hobgoblins. move horizontally across the field from left to The terror has a visual source before literary right, perhaps because we are habituated to sources, and its cause is how darkness looks. reading (therefore, to scanning) in that Absence of color differentiation is known to direction. disturb some people. Walter Cohen noted The absence of landmarks familiar in an ill- anxiety among subjects in tests that were uminated environment may explain the charac- "attempts to study the effects of uniform 50 Light of Day, Dark of Night

stimulation over the entire visual field," or Nocturnal inability to distinguish forms Ganzfeld (Cohen 1966, 306). The German term removes the comfort of the daytime means whole field or entire field. Typically, environment in which we orient ourselves to the subjects are exposed to dimly illuminated, accretions of space-time we have grown onecolor environments (the usual color is white) accustomed to interpreting as corporeal objects. in which no forms can be seen. Cohen's Goethe believed that confrontation with observers were asked to peer inside a structureless color, including the color of the translucent sphere that had its interior night, created a sense of deprivation. The color illuminated from the outside. "The most field of the darkness can suggest invasion by the representative description of the homogeneous inchoate but also respite from the confines of Ganzfeld is that of close, impenetrable fog. The structure. Potentially terrifying in one case, it experience is a unique one, and most 0's frees the spirit in the other. [observers] have difficulty in describing the Night is frightening, and its imagined field in terms usually associated with visual hobgoblins are memorialized in folk tales from phenomena" (Cohen 1966, 313). The fog was around the world. I regard the hobgoblins as a reported as "a hazy insipid yellow," or "misty, pictorial expression of the idea that a formless like being in a lemon pie" (Cohen 1966, 309). world of color is irrationally different from the Cohen's subjects found the experience world of everyday experience. The color unpleasant, and "anxiety and fear of blindness conditions of night are so different from those under somewhat similar conditions also were of daytime that the most bizarre encounters reported by Hochberg, Triebel and Seaman." seem possible. Seeing "nothing" is a Calling a field of color a fog implies that confrontation with the primitive Ur-vision, with objects ought to be present or are concealed what human beings see when ostensibly nothing by the "fog." What is expressed is resistance is to be seen. to the idea that we can see without seeing Brown is often identified as the most com- forms. Form is equated with meaning and monly occurring of all terrestrial colors. But structure and can suggest logic, an association black, including the entire range of nominal founded on the idea that forms can be blacks and near blacks, is observed more often. described in terms of geometry, and geometry is Black, whether absolute or nominal, is the said to be logical. color of night, including the expanses of The conception of form as a visual, unending night between the stars in the sky. intellectual, or psychological necessity is wrongheaded but commonplace. Its simplest expression is a demand for more than one color, Look Back Time familiar in responses to works of art. When first The uncertain nature of the relationship between exhibited, paintings by Barnett Newman and black (which is a color), darkness (a quality of others in which a canvas is covered with a black) and nothingness (a metaphysical single color were found unsettling by some concept) is touched on in the so-called look viewers. Form-seeking resistance to back time of contemporary cosmology and monochrome apparently also explains why astronomy. The idea of look back time is experiments in sensory deprivation disorient founded on the reasoning that light from the subjects untrained in meditation, and why quasars at the edge of the known universe, meditation is difficult to learn. The darkness unusually bright celestial objects at distances within can be as unsettling as the dark of the approaching fifteen billion light years, must night, evidently for the same reason. have begun traveling toward earth an enor-

Light of Day, Dark of Night 51 mously long time ago. What we see, for this crop up in several disciplines. If, say, the reason, must be interpreted as an image of what universe was originally a black nothing, what used to exist: of the way the universe looked in expanded or exploded to create the big bang? a distant past when galaxies began to form. The modern explanation that the expansion of Look back time tells us about the structure of the universe means the expansion of the spaces the universe but also about the nature of visual between the stars is not entirely informative. imagery. How did an original black nothing become the The quasars, according to one theory, are space that expanded? young galaxies, formed after a period of darkness that followed the big bang, the cataclysmic explosion by which the universe The Causes of Nothing was created. The quasars, if this theory is Cultural commitment to the concept of causality correct, probably no longer exist, except as leads us to believe that any colored image we images formed by the light and other radiant see has a cause. The causative something may energy they emitted billions of years ago. vary from light waves reflected by objects (as Beyond a limit estimated at fifteen billion light when I see a red apple) to human neu- years, the sky looks uniformly black and rological processes (as when I hallucina- contains no celestial objects. We are looking so torily see what is not there). Images that are far back into time (or into four-dimensional black in color fail to consistently conform to space-time) that there is nothing to see, other this paradigm. Seeing a black pair of shoes, like than the darkness that existed before the seeing a red pair of shoes, is agreed to be quasars. caused by something: likely, by the presence Why does empty space look dark or black, of shoes of that color. But the black of the sky rather than "looking" (if this is possible) or of the night is uniquely said to have no imperceivable? It is the familiar question of cause, which is what being caused by nothing why, in the story of the Creation, God's first implies. act is the creation of light (Genesis 1:3) but Unless either causality or nothingness can prior to that "darkness was upon the face of the be redefined, curious questions will continue to deep." Why does God create the world out of intrude. One is how an effect (black) can be darkness and not out of absolute nullity? created by an entity (nothing) presumably devoid of properties, including the property of Black, Nothing, and Look Back Time being able to cause an effect. Another is why I doubt that the question of the relationship, if nothing causes no effect in some cases but the any, between black and nothing is semantic as effect of black color in others. it arises in cosmological look back time. If If the black color of the sky cannot be blackness permeated the unpopulated attributed to nothing, it must be something. spacetime of the early universe, which had no Several possibilities occur for what that stars to radiate light, we need not assume that something might be. Word drifts back from nothing existed. In visual terms, the black color cosmological circles that a missing mass cannot of the darkness existed, as did the darkness, in be located in the universe, which appears to whatever form this formlessness possessed in include things we cannot see (Hawking 1988, isolation from its color. 45). Black, in the night or in the sky, may be Equating either with a vacuity leads to the color of a form of radiant energy antece- problems with words, prototypes for the dent to, or more primitive than, light: the semantic tangles about the color black that color, say, of space-time, or of the radio waves

52 Light of Day, Dark of Night

that surround us in space and are thought to be and objects do not have "normal" ways they remnants of the big bang. ought to look. The size and shape an object The enigma of black reduces to the ques- displays for a viewer depends on the interval of tions of what we mean by absence of light, and space between object and viewer and on the why the condition is assumed to imply an orientation of the object. For an architect at a absence of anything. Whatever the answers, the drafting board, a building can be drawn in status of black, in explanations of color, is a correct perspective from any of an infinite Gordian knot that defies attempts to untie it. number of points of view. No point of view is Part of the tangle is caused by the double more normal than any other. standard about the color that acculturation Human beings and animals recognize imposes. Some people say nothing can be seen objects seen in unfamiliar orientations, but in the dark. If asked to admit that the color of human beings both understand and resist the the darkness can be seen, they reply either that infinity of possibilities. We maintain mental this is not a color or that the blackness is what picture books in which each object is remem- they mean by nothing. Yet nobody seriously bered in the single aspect considered normal. believes that black is a definition of nothing Paleolithic painters depicted animals only in that can stand scrutiny. Black is the name of a side view, although they saw and recognized color. living animals from many perspectives. Egyp- tian and Sumerian artists combined "typical" views of different parts of the human body, Look Back Time as Perspective evolving composite figures; head and legs are Cosmological look back time, although not seen in side view although the chest is seen in customarily identified in this manner, is a four- front view. dimensional space-time perspective, a superset Today, thousands of years later, twentieth- of the more familiar perspective of three- century art students often have difficulty learn- dimensional space. Look back time is a visual ing to draw the human figure from unusual phenomenon. It looks as it does because of how points of view. The most difficult are orienta- we see. tions in which relative sizes of body parts are In terrestrial space, objects look progres- transposed. The student is able to see that in sively smaller as they recede into the distance. some circumstances the model's hand can look The phenomenon, called perspective, is rarely larger than his or her head. This is hard to recognized as an everyday example of the rela- accept because it offends the conception of tivity of either space or the extension of images what is normal. in two-dimensional visual space. But it pro- Perspective is a universal phenomenon vides a perfect illustration of what Einstein because nothing escapes from it. It has small concluded through mathematical, rather than effects as well as large. Roman painters, cou- visual, reasoning. No coordinate system is rageous enough to attempt its depiction (it was privileged, because the universe has no fixed largely ignored by the artists of earlier cul- center. We imagine ourselves at the center of tures), were reluctant to recognize its all- the universe, because practical reasons exist for pervasiveness. An anonymous painter from doing so. I need to know how far a door is Pompeii failed to realize that if two legs of a from me, though I know the door is not the table are the same length when measured by a same distance from other people. ruler, the leg farther from the viewer will look Absence of a privileged coordinate system shorter to that viewer. implies that perspective is a factor in vision, Although most people sense this and other Light of Dar, Dark of Night 53 similarly minute perspective effects, few have understood. The understanding intrudes in occasion to think about them. A Roman intimations of alienation or separation, or in the painting with the perspective inadvertently mythic urge to look into the future, feelings that reversed (with "wrong perspective" by modern reflect genuine insight into the nature of the standards) looks wrong to the modern viewer. physical world. The loneliness of God is the But many people cannot immediately identify loneliness of human beings. We dream through what disturbs them or how they would modify our nights alone and die alone. We also see the painting to make its perspective look alone and live alone, each subject to our own "right." personal technicolor visions in the privacy of Moving from tables to extraterrestrial our own unique coordinate system. spaces, the assumption inherent in the fourdimensional perspective of look back timewhich concerns time as well as astronomically large spaces-is that celestial Black as an Absolute objects we see at great distances are old, If anything causes a greater amount of because light from them has required time to conceptual difficulty than the habit of equating arrive. What we see today is the way these black, including the black of the sky or of the objects used to look, millions or billions of night, with nothing, it is the custom of regarding years ago. This idea, too, demands the color as an absolute. Most people find it easy generalization, because the earth and its objects to understand that a color name such as red (as the cosmological principle implies) cannot refers to a range of colors, rather than a single be materially or functionally different from the variety of shade. Vermilion, magenta, alizarin, rest of the universe. and crimson are names for types of red, Like rays of light from the quasars, rays subcategories within the larger category. Black reflected from the proverbial yellow chair in and white similarly refer to ranges of color, my room take time to travel to my eye. In although this is more difficult to grasp. Terms absolute terms the time is so brief I can ignore such as near black and off-white misleadingly it for practical purposes. An interval, imply that no shade of color is black or white nonetheless, must be assumed. It can neither be unless an extreme or absolute in that range. reduced to zero nor eliminated. That all black objects are not identical in The curious implication of this infinitesimal color (and therefore are not all absolute black) interval is that every color I see, or every set of can be easily demonstrated by collecting a dozen colors combined to form an image of an and comparing them. The difference between object, must be understood as an image of the tonalities in this range is played on in the black way the perceived object used to look. Arguing paintings of Ad Reinhardt. Each is a sixty-inch around this is impossible, even if "used to" square of canvas divided into nine (three by implies how the object looked billionths upon three) squares. Each of the nine embedded billionths of a nanosecond ago. I am subtly squares is painted a different shade of black. separated from the things that I see, because at None of the nine blacks is an exact color match the moment I see their images the objects are for any of the others. slightly older. Similar games, as easily played at the other I understand this to imply that all colored extreme of the value scale, are familiar images I see are those of objects from a dif- throughout the history of art. From Jan ferent time than my own. We can see only into Vermeer's Young Woman Standing at a the past although we live in the present. All Virginal (National Gallery, London) to Kazimir time is look back time, which is intuitively Malevich's White on White (Museum of Mod- 54 Light of Dar, Dark of Night

ern Art, New York), innumerable paintings play color. The coinage is unfortunate. on the point that all whites are not color matches for one another. Like black or any other color name, white is a label applied to a Black, Blue, and Sky range of thousands of closely similar colors. Among the chromatic colors, as Goethe and others have remarked, blue is the most similar to black, not just because it is dark. Achromaticity Generations of housewives have added bluing Words confuse about color as often as they to laundry because many people appear to elucidate because language is inherently believe that a white that is slightly blue looks limited or because human confusions are more genuinely white than one that is incorporated into it. In English, one locus of yellowish, grayish, reddish, greenish, brownish. difficulty about both black and white lies in the There is a visual affinity, in other words, double meaning of color. In its broader sense, between blueness and the color quality called the word names the class to which any and achromaticity, seen in black, gray, and white. every color belongs. In its narrower, color is a Speakers of some languages use a common synonym for chroma: the redness, blueness, name for blue and black, which led early greenness, or other hue quality that makes it anthropologists to assume no difference could possible to distinguish one spectral hue from be seen between the colors. Discounting that another. extravagant conjecture, a refined visual In thinking of black as color in one sense observation is involved. Blue and black look but not in another, we rarely reflect on the similar, by any ordinary visual criterion. At the oddity of using color to refer, at one and the dark end of the value scale, navy blue and black same time, both to some colors (the spectral are more easily confused than any other two colors) but also to any and all colors. To colors, particularly at a distance. At the pale stabilize the ambiguous position of black, end, relationships are more ambiguous. Light white, and gray (the color obtained by mixing gray looks more similar to or is more easily black and white), the term achromatic colors confused with a light blue of similar value than came to be applied. A syntactical peculiarity in a similarly pale pink or yellow. But light violet itself, achromatic color perpetuates the idea may be more similar in color to light gray than that black, white, and gray are colors without light blue is.

CHAPTER 7 Light and Dark in Perspective

Sight being, as I conceive, in the eyes, and he who has eyes wanting to see; color being also present in them, still unless there be a third nature specially adapted to the purpose, the owner of the eyes will see nothing and the colors will be invisible. Of what nature are you speaking? Of that which you term light, I replied.

Plato, The Republic

olor, unlike form, cannot be reduced to Something and Nothing geometry, which is able to present Because the spaces enclosed by their perimeters in visual terms ideas that remain are treated as if divorced from those outside, the abstract if expressed arithmetically. Plane shapes of Euclidean geometry float in a sea of geometry is a primitive geography of nothingness. We imagine the stars in the sky in twodimensional universes. Supplemented by a similar manner, white objects adrift in empty the laws of perspective, it introduces the space. Art contributes further images of topography of the picture plane and the visual something displayed in a field of nothing, a field. But theorems memorized in high school primitive assessment of the nature of visual are not always consistent with perceptual phenomenology. Most painters before the experience. Those theorems convey rules about Romans drew human and animal figures against forms drawn or projected on planar surfaces. blank backgrounds. Black curtainlike expanses They rarely acknowledge backgrounds, the are suspended behind the sitters in negative spaces surrounding forms. Plane fifteenth-century Netherlandish portrait geometry is as silent about the nature of its paintings. Like the blank backgrounds, the black implied universe - a picture plane - as it is curtains deny the continuousness of two- about the origin of the Euclidean point. dimensional pictorial space by asserting a pri-

56 Light and Dark in Perspective

vate dimension of their own. The interesting visual and conceptual aspect Although the inconsistency is not always of this computation is the elusiveness of zero, a easy to grasp, all such vignettes deviate from metaphorical link between color and number. visual experience. They express the duality Just as no amount of lowering of light can that Ivins called the tactile-muscular ambience reduce vision to nothing (we always see the (Ivins 1946). Unlike the objects I touch, those color of the darkness), an infinite number of I see, whether red triangles or white chickens, repetitions of the process of dividing will not are not located in emptiness. Each is sur- reduce a numerical quantity to zero. In rounded by an environment continuous with subtraction, where zero can be arrived at, it itself, similarly reducible to a simple or proves to be a bridge rather than an abyss: the complex pattern of color spots. We acquire or point of transition between the positive and fortify incomplete conceptions of two- negative numbers. dimensional universes when we study geom.- If the zero or cipher has a counterpart in the etry. The Greeks sing siren songs from their two-dimensional universe, I prefer to imagine it graves, pursuing us with their unsophisticated as the vanishing point of a perspective system, conviction that tactile experience is the model which, like a Euclidean point, can be located but for all experience. never perceived. No visual analogue exists for As in the popular idiom, "a nothing per- silence, because seeing implies seeing color, son," an object labeled nothing is summarily and we never see nothing. In art, the concept of dismissed from attention, the subjective or negative space, the spaces around forms, operational meaning of an otherwise incom- became important during the late nineteenth and prehensible term. Euclidean geometry ignores early twentieth centuries. An understanding of the spaces around forms. And cosmologists negative space makes it possible to deal with a and astronomers were not originally con- two-dimensional matrix, say, the picture plane cerned with studying the blackness or noth- of a painting, without the preconception that ingness between stars. Interest centered on some parts of the composition are important and investigations of any something, whether cos- others are not. mic rays, radio waves, or light waves, possi- Reasoning based on tactile rather than visual bly or actually traversing the void. The recent experience remains a more widespread kind of bubble theory of the nature of the universe is reasoning. It begins with eliminating what is exceptional in this regard. Its foundation is the nothing, not significant. The job is to discover long-neglected negative space: the shape of the what is significant. In visual thinking, interstellar blackness as we perceive it from everything is significant, a paradigm based on earth. the experience of seeing color. For the Although usually less visual than geometry, functioning eye, color is all-pervasive. It arithmetic rises to finer metaphor in this case. permeates all parts of the visual field at all The cipher or zero, invented by medieval times. Hindu or Arab mathematicians, expresses the I am not surprised that in mathematics the idea that an entity cannot be ignored just cipher or zero was a non-Western invention. We because nothing is the label applied to it. By rarely reason effectively about anything, acknowledging a difference between 15 and including color, that we regard as formless or 105-by recognizing that nothing does not inchoate, as unimportant or as nothing. I find mean no effect-the cipher enabled modern some students unable to grasp the concept of computation. negative space. They list the names of objects Light and Dark in Perspective 57 they see in paintings and insist, against all What any human being sees at any given reason, that nothing else is included by the instant is a portion of time and space, limited artist. Often such a student is highly intelligent by the spatio-temporal perspective of that with an overdeveloped conscience, a person observer. Comparable parameters govern who "has standards" and takes society's representation in those visual arts that are imperatives seriously. Separating what will be twodimensional. To represent, or depict, a looked at from what will be ignored is a habit moment of time and space in a painting, I do that cannot be transcended. not use time and space. What are manipulated Visual thinking is sometimes called creative are the colors of paints, again, to create a thinking, as if there were any purpose to perspective. thinking that is not creative. I think of the visual To assess visual experience and the place of paradigm as more than an aesthetic nicety color in this experience, we need to consider all useful in understanding the visual arts. Most of that we see at night, not just how objects look in our better ideas, are based on a visual, not the daytime. Astronomy and cosmology have a tactile, paradigm. great deal to teach us about color and in fact are Visual thinking is thinking for the present. studies of colors, although we rarely try to No other kind of thinking releases us from integrate findings in these disciplines into our unthinking recycling of the misconceptions of understanding of color in general. As Hawking earlier peoples. In tactile sorting of the noted, for the vast majority of stars, "there is significant from the insignificant, where will the only one characteristic feature we can standards come from? We inherit them from the observe-the color of their light" (Hawking 1988, past, condemning ourselves to live out our lives 37). The data of astronomy and cosmology as if we had been born a hundred or a thousand (except radio waves) are acquired largely from years ago. A great many misconceptions and observation of the varicolored image we call the poorly thought out conceptions about color are sky and the colored spectra formed by light traditional. They should be subjected, one by emitted from distant stars. The one, to a scrutiny that is visual, that does not much-popularized red shift is an observed color allow the possibility that vast areas of phenomenon. It provides our sole evidence for phenomenological experience can be the expansion of the universe and, beyond that, unimportant, insignificant, nothing. for the big bang theory, which holds that the origin of the universe was a cataclysmic explosion. Color in Perspective T. S. Eliot penned marvelously biting words Color might be defined as the mode in which when he heard that the universe was expand- time and space appear within the confines of ing. He wondered how anyone knew. But an a planar or two-dimensional universe. More expanding universe is the most coherent expla- than one color exists. We see an array of spots nation, perhaps the only conceivable explan- of different colors, an ambience often ation, for why the spectra of light from distant explained by saying that we see color and form. stars is shifted toward the red end of the visi- But forms cannot be seen at night. And forms or ble light spectrum. The red shift increases with shapes, in a two-dimensional matrix, always distance, implying a limit beyond which no reduce to arrays of color spots. I prefer to visible light can reach us from stars. This is not imagine that the basic visual elements are because the stars are so far away, but because color and perspective, or colors in perspective. they are moving away from us so rapidly. The 58 Light and Dark in Perspective

edge of the visible universe is reasonably spun yarns about how, say, the mythological defined as a color phenomenon: the distance at giant Orion came to be placed in the sky as a which the red shift becomes so extreme that constellation of stars. Later the constellations light from more distant stars cannot reach us. were thought of as guides for navigators. But The tactile habit of thinking in terms of the pictures in the sky were not originally polarities gives a gratuitously violent theory of picked out to help navigators, and the navigators how the universe began, though the evidence did not need the stories. Possibly the idea of presented would fit more modest theories as drawing pictures on surfaces was inspired by well. If all celestial bodies are moving apart the'game of drawing pictures in the sky by from one another (or if the spaces between the imagining lines linking the dots of the stars. stars are expanding), this need not imply that Today the stars are viewed with more the process began with a cataclysmic sobriety. The modern understanding is that the explosion. The beginning might as easily have images in the sky are not images, not resembled a pebble dropping unnoticed into chimerical. They are neither just colors, water or, as in recent theory, the foaming of a pictures, nor essentially visual phenomena. The silent sea of black bubbles. spots and dots seen through telescopes are The celestial bubbles, discovered in mapping accounted for by celestial objects believed to of a large sector of the sky, consist of starless be "there." We may pass lightly over the expanses of approximately ovoid shape. The embarrassing question of whether the blackness edges of the ovoid shapes are marked by the surrounding these objects is similarly there. galaxies squeezed between them. Whether or The assumption that it is remains difficult to not it proves out, the bubble theory is genuinely reconcile with the status of interstellar visual. It takes account of the entire sky, rather blackness as nullity: as an image of, so to than assuming that stars are to be taken speak, what is not there. seriously as something and the black spaces Meteorites fall to the earth, and rockets can between them to be disregarded as nothing. The be dispatched to the planets, so the question of bubble theory does not assume violence-a war whether objects in the solar system (by of opposites-where no evidence of violence is to extension, other celestial objects) are verifiably be seen. there appears to be laid to rest. The quasars, Cosmologists are unlikely to classify however, tell a different story and represent a themselves as investigators of colors. And those different class of visual image. Afloat in their astronomers, say, Fred Hoyle, who stress the four-dimensional space-time perspective, the importance of observation of the stars to the quasars were there at one time, a time development of human thought rarely mention irreconcilable with our own. As that we could not observe the stars if we had cinematography similarly suggests, the bare hands and not eyes. Nor do they mention that a fact that we collectively see something and can race of color-blind human beings would have take photographs of it does not prove it is there. had no way to see the red shift. Astonishingly, objects-and not, as we Long before people observed the stars with imagine, images-may be the more ephemeral scientific detachment, they used them as a class of entity. The images of quasars seen form of celestial entertainment. Perhaps they through telescopes today have endured for fif- stared at the sky with the fascination modern teen billion years. This life span is longer than children give to television screens. People that enjoyed by the quasars in the form in imagined one dot joined to the next to form which we presently see them. In visual terms, the pictures we now call constellations. They the most interesting aspect of the images of Light and Dark in Perspective 59 quasars is the failure of correspondence not just more blue-gray, than those that are between image and object. If we could nearer. They also display a lesser degree of instantaneously transport ourselves to the spot either color contrast or tonal (light-dark) where we see the quasar, it evidently would not contrast. These effects are not explained by the be there. This failure of correspondence has a presumed blueness of air. Nor can they be visual parallel in the more limited world of replicated by looking through blue-tinted glass. terrestrial perspective. A more problematical area is that if aerial Just as seeing a quasar does not prove it is perspective is caused by the blueness of air, it there, seeing an elephant progressively diminish is not, strictly speaking, an effect attributable to as it walks off into the distance does not prove distance. Indeed, the color effect may be that the elephant is shrinking. That objects look accidental. We are left to assume that on a smaller when farther away has been pointed planet where the atmosphere is slightly orange, to-incorrectly, I think-as evidence of the distance objects would look orange rather than unreliability of the senses. Perspective effects, blue. in elephants or quasars, are visual evidence of Whether or not aerial perspective should be the absence of any privileged coordinate regarded as a perspectival effect, a modification system: of the relativity of space and time, or of caused by distance, it has no parallel among the the role of vantage point in determining how the colors of extraterrestrial spaces. The moon and world looks. More important than whether the stars look more white than blue-gray, though walking elephant is shrinking in any absolute the same atmosphere intervenes between them sense is that the elephant is not shrinking from and us as between any viewer and distant its point of view, however tiny it may appear to scenery. Because of the red shift, light from have become to a conscientious observer. distant galaxies looks progressively more red rather than more blue. Only if the universe were Time, Perspective, and the Colors of shrinking, theory implies, could this red shift be Objects replaced by a violet shift or a blue shift. If an evanescent panorama of perspectival Is the red shift itself a color-perspective effects is what we see when looking out into effect? If caused by the speed at which distant space and time and these effects reduce to galaxies move away from us, and not by their arrangements of color spots, are colors subject absolute distance, one might say no: the color of to what are loosely known as the laws of light received from distant stars does not perspective? On earth, distant objects look depend on their distance. The argument could bluish or blue-gray and more pale than nearby be yes on another basis. Distant galaxies move objects. The poorly explained effect has long away more rapidly than those that are near and been known as aerial perspective. Although are more red-shifted. Speed and distance in this attributed by Leonardo da Vinci and others to case are inseparable. the effect of air, which is said to look blue in A crude perspective of color value (light- thick layers, the explanation is not entirely ness or darkness) can be identified in satisfactory. astronomy and cosmology. If two identical Beyond the effect being explained, no cor- stars are at different distances, the ratio of their roborating evidence is available that air has a brightness (a color relationship) is said to be slightly bluish color or that the blueness is in inverse proportion to the squares of their proportional to the thickness of the layer. Fur- distances. This rule for the perspective of vis- thermore, distant objects look more pale, and ual brightness, of little interest except to

60 Light and Dark in Perspective

astronomers, parallels the more familiar rules The sun is a G star, Sirius an A star, and for the perspective of visual size. Beyond this, Betelgeuse an M star on this scale. An oddity of the colors of radiant celestial objects too close the scale is its inconsistency with the scale of to be strongly affected by the red shift reveal of visible light. The hottest stars are little about the distances of these objects. Color, not, so to speak, white-hot. They are not reddish, however, conveys a great deal of other the spectral extreme that extends into infrared. information. The chemical composition of stars Nor are they bluish or violet (the other extreme). can be determined from their spectra. And The green cast of the hottest stars suggests a color, as Leonardo da Vinci and his preponderance of radiation median between red contemporaries recognized, is a more reliable and violet, the extremes of the electromagnetic measure of the absolute surface temperature of scale for visible light. celestial objects than either their brightness Within terrestrial limits, as in the sky, a when observed from earth or tactile sensations simple increase in space interval has little effect of warmth they induce. on color but modifies form considerably. A Human beings reason by analogy, and often nearby elephant, compared to another farther the better analogy incorporates more keen away, looks more different in size than in color. observation about the phenomenological world. A perspective of colors in time can be identified, Thus, Leonardo complained that, "some say however. I mean by this that the colors of those that the sun is not hot because it is not the objects in the world that reflect light are colour of fire but is much paler and clearer. To unendingly changing rather than fixed. Attributed these we may reply that when liquified bronze to variations in illumination, the effect implies is at its maximum of it most resembles the changes within time, the dimension required to sun in colour and when it is less hot it has more allow illumination to vary. of the colour of fire" (Leonardo 1939, 281). Examples of the phenomenon suggest that Leonardo would not have been surprised at one objects have no real colors, because no fixed classification system used in astronomy today. point can be located in the evanescent pano- It ranks stars according to their surface rama. What color is, say, a red rubber ball illu- temperature, estimated from their spectra or minated by purple light? Questions of this type colors (Gallant 1961). elude answer, other than that the color of any object varies in its color according to changes in lighting conditions over an interval of time. Ranking of Color Surface Under a blue light, a green light, a purple Star Temperature light, the colors in an environment change, O greenish approaching the hue of the illumination. The white 35,000° C+ phenomenon is not limited to circumstances B bluish 15,000-35,000° C in which the variation in light is extreme. A white 7,500-11,000° C Colors that appear to match under the yellow- F yellowish 6,000-7,500° C G yellow 5,100-6,000° C ish cast of artificial illumination often look dis- K orange 3,600-5,100° C similar or are discovered to be mismatches M orange-red 3,000-3,500° C when viewed under outdoor light. Experience R orange-red c. 2,300° C teaches the folk wisdom that color matching N orange-red c. 2,600° C ought to be done in daylight. S orange-red c. 2,600° C Light from the sun changes in color qual- ity throughout the day. The blue cast of early

Light and Dark in Perspective 61 morning and evening veils the environment in nary electric light bulbs shed a yellowish light, that tonality, causing red objects to look bluish. becoming more yellowish with age. The at noon is golden, although not so 3,400°K photoflood bulbs and 3,200°K tung- yellowish as tungsten light. Further variations, sten bulbs used in photography similarly give many studied by the French Impressionist light that varies in color according to the age of painters, run according to seasonal cycle, the bulb, a matter of concern in cinematog- latitude and longitude (Mediterranean sunlight raphy. The color of these bulbs is also affected differs from that of the Arctic), or the transient by spikes and surges in electric voltage. weather conditions that provide both gray days Of all commonly available types of bulbs, and golden days. quartz lights show the smallest degree of color This dizzying continuum of permutations variability. The color of an object might be presents a barrier to identifying the colors of defined as the color the object exhibits when objects. A particular red object may look one illuminated by a specified number of quartz color in daylight, a different shade under artifi- lamps, of specified wattages, placed at speci- cial light, another under blue light, yet another fied distances. Quartz bulbs are expensive, the under green. The object apparently has no real reason they are not used for all-purpose illu- color-or no constant color in isolation from its mination. The costliness would limit any stand- environment. The colors of objects, however, ard based on their use. The inch and the are not entirely dependent on illumination, centimeter are successful standards because which speaks against the possibility of rulers can be produced cheaply and distributed explaining them solely in terms of the light widely. A large ruler-using population has falling on them. There is no light under which therefore learned to measure in inches and feet black velvet can be made to look chrome or centimeters and meters. The same popula- yellow. tion cannot be expected to accustom itself to thinking of the colors objects present when viewed by quartz lights. The lamps are too Developing Color Standards expensive to become as widely available as Newton's belief that colors reside in light, and rulers. not on the surfaces of objects, must have been The ambiguity built into color names has a inspired by the chimerical quality of surface positive aspect in this case. As Moneys paint- colors, a scintillation with an extraterrestrial ings of the facade of Rouen cathedral show, counterpart in the twinkling of the stars in the with considerable poetic license, an object sky. We cannot even assume that a color varies in color over the course of a day. A red swatch being used for comparison will invari- apple lying in a field is not the same red from ably look the same color irrespective of the morning to evening. Because illumination is illumination. How can stability be imposed on not perfectly even, the apple is also not uni- this flux? Can the task be accomplished? form in color over its entire surface at any sin- If normal or standard lighting conditions gle moment. But the color name red is so loose were established, the real colors of objects and covers such a wide range of color varia- might be defined as those exhibited under tions that we may safely conclude the apple those normal conditions. Accomplishing this is red. The greater number of color variations requires seeking out the type of illumination on the apple's surface will probably fall in that that is most invariant in color. The task class. Red names a range of colors, not a sin-- presents considerable practical problems. Sun- gle shade or variety of the color. But a red light is too changeable for the purpose. Ordi- object is many shades of red, a nice matching 62 Light and Dark in Perspective

between color names and the colors of the ivory black pigment lies in the blue range. The objects to which they refer. pigment looks black rather than blue because so little blue light is reflected. Some blacks have a brownish cast rather than bluish. But, so far as I Colors Created by Light know, no browns in the animal world, where Color is often thought of as a skin or surface the color is common, are created by a on objects. But natural light phenomena mechanism similar to that which accounts for include effects only loosely associated, if the Tyndall blues. associated at all, with the surfaces of objects. These effects include the colors of oil slicks and (or those of the sky and the Rainbows night). Colors in oil slicks and soap bubbles are Among natural light phenomena, the rainbow given their most convincing explanation as and related forms of the solar spectrum are interference patterns. These patterns are unique in presenting colors in an invariant created when different wavelengths of light, order. We do not know when the discovery was some reflected from the upper surface of the first made that artificial "rainbows" could be film and some from the lower, impinge on one created. But, as Koyre pointed out, study of all another. manner of spectra, including those formed by The Tyndall blues, explained in terms of prisms, began in antiquity. It continued diffraction, account for many of the blue colors throughout the Middle Ages and the seen in beetles, butterflies, birds, and Renaissance, proliferating during the mammals. Named after the Irish physicist John seventeenth century (Koyre 1965, 3).' Leonardo Tyndall, who discussed the phenomenon in da Vinci advised the seeker of spectra to 1869, they provide another example of the examine "the roots of radishes which have been ubiquitous visual affinity between blue and kept a long time at the bottom of wells or other black. The Tyndall blues might as stagnant water." In the water, each radish root is appropriately have been called the Tyndall "surrounded by a sequence of colours like those blacks. They are typically created when a layer of the rainbow" (Leonardo 1939, 26). of melanin, a black pigment, is overlaid by Prior to Newton, spectra, whether seen translucent ridges, scales, or other structures around radish roots or through prisms, were not that scatter light. A blue wing feather from a regarded as central to any general theory of blue jay, for example, contains no blue color. Robert Boyle (1627-91), with more pigment. When crushed, the feather is reduced typically scattered reasoning, conjectured that to a black, not blue, powder. white animals might be more common in cold Why the scattering of light over a black pig- climates, presumably because the temperate or ment ought to look blue rather than, say, red, tropical sun "brought out" more colors. In is suggested by a chart of reflectances of artists' Experiments and Considerations Touching pigments compiled by the Fogg Art Museum Colours, Boyle listed six major and additional (see figure 23-4). Black is called a reflector of minor hypotheses current among his no light, too loose an assessment. Any black contemporaries (Boyle [ 1664] 1964, 85). other than absolute black would reflect some One theory, attributed to Aristotle, "de- light, raising the question of what color or rives Colours from the Mixture of Light and colors this feebly reflected light would be. The Darkness," a recycling of the belief that the Fogg Art Museum compilations indicate that world can be explained as an interplay of the most profusely reflected by opposites. A second theory, accounting for Light and Dark in perspective 63 colors in terms of a sulphurous principle in experiments, colors are created by "an internal bodies, had been borrowed from alchemy. In light of the more lucid parts of the object" the updated version current in Boyle's day, the (Boyle [1664] 1964, 84). A variation of the colors of objects are explained by their relative internal light hypothesis held that color, as if a proportions of salt, sulphur, or mercury. permeating miasma, is "a corporeal Efflu A third proposal is that colors are caused, as vium issuing out of the Colour'd Body" (Boyle Plato suggested, by corpuscles emitted by [1664] 1964, 85). Robert Hooke (1635-1703), bodies. The model of the universe developed by who began his career as Boyle's assistant, Rene Descartes (1596-1650) follows this proposed that "colour is nothing but the format. Revolving globules of etherialmatter disturbance of light by the communication of produce different colors according to the speed the pulse to other transparent mediums, that is, of their whirling (Principia 3:52). by the refraction thereof " (Whittaker [1910] Rapidly spinning particles, Descartes 1951, 1:16). declared, give a sensation of red. Slower ones Hooke, who later was to argue bitterly with produce yellow; the slowest account for green Newton, declared in his address to the Royal and blue. Why did Descartes believe the most Society (1671) that "blue is an impression on rapidly spinning articles would be red? He must the Retina of an oblique and confus'd pulse of have been inspired by the traditional symbolic light, whose weakest part precedes, and whose association of red with blood, fire, , and strongest part follows . . . . Red is an impression states of high energy. Green and blue tend to be on the Retina of an oblique and confus'd pulse associated with spirituality, peacefulness, or of light, whose strongest part precedes, and passivity. whose weakest follows." In other theories that predate Newton's

CHAPTER 8 Newton

Even Galileo himself was not the perfect scientific man. Perfection was reached only in the person of Isaac Newton. J. W. N. Sullivan, The Limitations of Science

oday the proposition that color depends on the spectrum cast on his wall. Newton's on light has accumulated ornate conclusions were initially published in "A accretions. The idea is rarely defended Letter of Mr. Isaac Newton . . . Containing his (although it might be defended) by observing New Theory about Light and Colors." The letter that at night, or in the absence of light, color is catalyzed a bitter controversy with Robert limited to the color of the darkness. Daylight or Hooke, recorded in an extended correspondence some other form of illumination is required to preserved in Philosophical Transactions of The see other colors or to see more than just one Royal Society (80:3075). A final exposition is color. Possibly, "light causes color" means no included in Opticks (1704), which Newton more than that. But we give greater weight to published more than three decades after his scientific explanations that to everyday original experiments. understandings. The line of reasoning regarded as more authoritative begins with the experiments of Sir Isaac Newton (1642-1727), The Inherent and the Phenomenological Urfather of modern color theory. Disagreement between Hooke and Newton In 1666 Newton, at the time a university turned on Hooke's belief that colors are caused undergraduate, passed sunlight through a glass when light is disturbed by refraction: the bend- prism. He proposed a theory of colors based ing of rays that occurs when light passes from

64

Newton 65 one transparent medium to another of differ- other of these manifestations is more normal, ing density, say, air to water or air to glass. real, or consequential is arbitrary. No basis Color, like the heat generated by friction (or exists for assuming that an object's real (or like the colors of bruises), was to be explained verifiable) state for a given set of conditions as the incidental by - product of a physical ought to be constant for all others. event. Newton demonstrated that light undergoes Hooke's theory can be construed to imply no irreversible change in passing through that glass prisms create solar spectra by prisms. But Hooke may have been equally inflicting damage, or irreversible change, on correct in insisting the light was changed, the light passing through them. Newton proved although not irreversibly. During the twentieth no irreversible transformation occurs. In an century, Whittaker and others were to take up experiment admired for its simplicity, he the argument that spectra cannot be seen (or do directed the spectral rays through a second not exist) until light passes through a prism or prism that was inverted. They recombined to other refracting device. The prism, in effect, produce a beam of white light. Newton manufactures the colors. theorized that white light is not changed into Hooke's argument (or Whittaker's) also colored rays by prisms. It is, instead, inherently implies that colors are not made manifest by colored. The proposition was subjected to more light. Instead, the sun, by radiating light, sophisticated scrutiny centuries after Newton's manufactures the colors we see by day. The death (Whittaker [1910] 1951, 1:14). So was remaining puzzle is what manufactures the Newton's famous remark that "the Rays to color of the dark of the night. One possibility is speak properly are not coloured" (Newton human consciousness, an answer impossible to [1704] 1952, 124). integrate with conventional understandings of Newton's criticism of Hooke followed the nature of the physical world. Another Hooke's criticism of Descartes. Hooke answer, no easier, is that the dark of the night is regarded light as a motion. Descartes called it a a priori, a color that exists without being tendency to motion. What is the difference manufactured in any ordinary sense of the between white light that is actually the white word. light it appears to be (Hooke) and white light Whether the colors of the spectral rays are that is inherently colored (Newton)? Because inherent in light or arise through its Newton's light is not colored but "inherently" modification is a seventeenth-century variation colored (it might become colored at some on that question human beings never tire of future time), the distinction is so exquisitely asking: whether colors are real or illusory, fine as to be semantic. When the inherent fails enduring objects or transient effects. One way to make itself manifest, it remains non- of putting the question to the test is to procure a phenomenological. set of real (nonillusory) colors. If the two sets No such thing exists as an object-in this of colors can be compared, we should be able case, a light beam-that looks one color but is to determine whether the colors we see look really another. Statements that seem to say different from the colors in the other set, the set otherwise characteristically mean that the of real colors. object shows one set of colors under certain Unfortunately, no set of colors is available conditions but another set of colors under other than those we see. The imaginary test other conditions. Light is white if not passed and the question it proposes to answer are though a prism, but it exhibits the spectral each equally misguided. The sense of the ques- colors after passing. Arguing that one or the tion is not improved by asking whether colors 66 Newton

are actually there, as opposed to just cated. Newton reports that he mixed blue and perceivably (ephemerally) there. Can we yellow spectral rays to produce green (some determine whether a color is actually there by commentators have questioned whether this is looking carefully at the color? I think not. possible) but found the green dissimilar in hue Sorting criteria-rules of language-are involved. to the singular ray of spectral green. No visual criterion can be identified that shows Rays could be mixed and unmixed at will. A the difference between a color that is there and test for whether a colored beam was one that just appears as if it were. Color cannot monochromatic was to pass it through a prism. be shown to be either there or not there and If it were an amalgamation of beams, these shares this characteristic with virtually would separate. If the ray was light of only one everything else. color, it would emerge from the prism a sin- Language, before it affects perception, gle beam. affects scientific theorizing about what we see. Newton was courageous enough to take a position in a difficult area. He found that The Nature of Light "Colours are not Qualifications of light derived The exchange of letters between Newton and from Refractions, or Reflections of natural Hooke is an episode in an extended controversy Bodies (as 'tis generally believed) but Original about the nature of light. The underlying and connate properties, which in divers Rays questions, originally raised bY the pre-Socratic are divers. Some Rays are disposed to exhibit a philosophers, became so closely entangled in red colour and no other: some a yellow and no color theory they will probably never be other, some a green and no other, and so of the extricated. The questions concern whether light rest. Nor are there only Rays proper and is corporeal and the relationship between light particular to the more eminent colours, but even (radiant energy) and matter. to all their intermediate gradations" (Phil. Two opposing views became classical and Trans., 6 [19 February 1671-72]: 3075). were repeatedly recycled. The first, in its earli- Newton's reasoning is a modified version of est formulation, assesses light as a stream of a theory Boyle traced to ancient Greece, where particles or corpuscles emitted by incandes- "the Peripatetic Schools, although they dispute cent bodies. The counterargument focuses amongst themselves divers particulars squarely on the issue of substantiality. If light concerning Colours . . . seem Unanimously consists of corporeal particles, it was argued, enough to Agree, that Colours are Inherent and these would collide when beams of light Real Qualities, which the light doth but crossed. But if the environmental scale of the Disclose, and not concurr to Produce" (Boyle particles were exceedingly small, the probability [1664] 1964, 84). A critical difference is that of collision might not be statistically Newton regarded colors as properties of light, consequential, and its occurrence might not be rather than of colored objects. detectable by instruments. The orbits of comets When Newton tried to analyze the spectral cross those of planets without collision rays further, he found this could not be accom- necessarily occurring. And the chance for plished. Each ray seemed elemental. Passing collision of molecules need not be assumed to the ray through additional prisms had no effect be significantly increased when one stream of on its color. Neither could the colors of the water intersects another. Thus, a mere absence rays be changed by other means.' Rays of of evidence of collision is not sufficient to some colors showed another type of unique- disprove corporeal particles. ness, in that their colors could not be repli- The classical alternate to the particle Newton 67 hypothesis is that light is a wave phenomenon. of what came to be called the Copenhagen The main counterargument is that waves, say, interpretation of quantum theory. The wave and those in water or air, require a medium through corpuscular descriptions were held to be which they travel. Light, which crosses a complementary ways of explaining light, both vacuum when it arrives from the stars, needed to understand the phenomenon apparently transcends the need for a corporeal (complementarity).z But (uncertainty) no conducting medium. experiment can demonstrate both at the same To account for propagation in a vacuum, the time ( Jenkins and White, 1957, 617). wave theory was shored up by proposing that More exactly, the uncertainty principle sug- space might be permeated by a gests the experimenter will be equally success- wavetransmitting vehicle designated the aether. ful in finding whichever configuration the Part of the reason for hypothesizing the aether experiment is designed to display. The con- was that empty space (incorporeality) was mcept has simple visual parallels. In one class assumed to be a vacant receptacle devoid of of optical illusion, ambiguous cubes look alter- physical characteristics. Yet vacuums had been nately concave and convex. By focusing atten- known since ancient times, and today their tion, the viewer can will which way the cubes characteristics are regularly exploited in ought to look. But the cubes cannot be seen as vacuum pumps and similar devices. both concave and convex at the same time. The aether was never isolated, in the sense At very high frequencies, say, in the X-ray that no light-conducting medium was proved range, radiant energy behaves as if corpuscles, to permeate vacuums or to be separable from but at low frequencies, as if a wave. Because them. The usual conclusion, that nothing exists wave properties predominate in the visible light in a vacuum (the aether could not be found), sector, scientific explanations of color generally is questionable on two counts. Empirically, we rely on reference to wavelengths of light, know interstellar space is not absolutely measured in either millimicrons (nanometers) or empty, a finding at odds with the traditional angstrom units. If radiant energy is to be wisdom that the universe can be divided into described in terms of the corpuscular theory in empty and nonempty parts or that the differ- its modern form, it is assumed to consist of ence between emptiness and nonemptiness is massless particles called photons, which belong significant. Conceptually, the absence of to a larger class of particles called bosons. etheriality "in" vacuums implies that the Photons are ejected from atoms in aggregates etherial element being sought, the medium that called quanta (hence, quantum theory). The conducts light, must be the vacuum itself. The different colors of light are correlated with corporeal (matter) and the incorporeal (the different energy levels in the photons. In the vacuum) are not necessarily different in every visible light sector, those responsible for red respect. Each is capable of transmitting waves, have about half the energy of those that account if light is a wave phenomenon. Each also can for blue. be transparent and, if so, will allow light to The varying energy levels of the photons pass. suggest Descartes's whirling globules, moving The classical argument about whether light particles colored according to their energy is particle or wave was never resolved. An levels. A difference is that each of the Carte- uneasy reconciliation was reached through sian particles revolves on its axis. The more Bohr's complementarity principle (1928). Sup- intricate choreography of the photon, shown plemented by Heisenberg's uncertainty prin- arbitrarily traveling in a straight line in the ciple, complementarity became a cornerstone Feynman diagrams used in quantum 68 Newton

mechanics, allows it to transcend time (it can either field. For Planck, the remaining task was move backward as well as forward in time) and to amalgamate mechanics into the system to be transformed into an electron or a positron. (Planck n.d., 145). These peregrinations through time cast doubt on In addition to inventing the Maxwell discs whether the relationship between light and color used for many years to demonstrate color is as simple as is usually assumed, and on mixing and devising the Maxwell triangle still whether light moves in the ordinary sense of the popular in adapted form in colorimetric studies, word. Maxwell left about a dozen papers on color. Newton denied Hooke's accusation that he Although he is said to have considered his was a proponent of the doctrine of the papers on color the most important of his corporeality of light. Any incandescent body, works, his reinterpretation of the light wave as Newton argued, emits corpuscles, which excite an electromagnetic phenomenon brought the waves in the aether, a reconciliation of sorts study of color in physics to an abrupt halt. between the wave and corpuscular theories. Some questions were never asked again because a definitive answer was thought to have been found. Others were relegated to psychology and If by any means those [aether vibrations] other life sciences, or to philosophy and of unequal bigness be separated from one aesthetics. From the seventeenth century until another, the largest beget a Sensation of a the twentieth, the word color is used with Red colour, the least or shortest of a deep precipitously declining frequency in the Violet, and the intermediate ones, of literature of the physical sciences. Russell intermediate colours; much after the complained, forty years ago, that it could be manner that bodies, according to their avoided entirely. The result was a gradual several sizes, shapes, and motions, excite curtailing, over the centuries, of questions that vibrations in the Air of various bignesses, legitimately could be asked. which, according to those bignesses, During the seventeenth century Robert make several Tones in Sound (Phil. Boyle had no hesitation about inquiring why Trans. 7:5088). and how red looks different from green (Boyle [ 1664] 1964, 91). The query lies far afield After Newton, rearticulations of the wave from the physical sciences today. Or it might theory were presented by Christiaan Huygens be disposed of by the modern scientific answer (1629-95), by the English physician Thomas that red differs from green because each Young (1773-1829), and by Augustin Fresnel derives from a different sort of radiant energy: (1788-1827). (1831-79) in classical theory, from light with a different reinterpreted the light wave as a form of elec- wavelength; in quantum theory, either from a tromagnetic wave, related to X rays, radio probability wave with a measurably different waves, and . Max Planck wavelength or from photons with a different (1858-1947), discoverer of the quantizing of energy level. energy, provided one answer to the question The modern answer, however framed, is of why Maxwell's theory found favor. It had circular, because colors are defined in physics an immediate utility. Irrespective of whether in terms of radiant energy. If the wavelength electromagnetic theory was true in any abso- of red light is agreed to differ from that of lute sense, it allowed optics and elec- green, this does not speak to the question of trodynamics to be connected. The single set why this should cause the two colors to look of laws it generated explained phenomena in different. Nor does it explain why they look Newton 69

different in the way they do. The visual question is similarly and destruction. The electromagnetic force, carried by the evaded by recourse to other disciplines that give attention, photon (the particle still thought to account for color and say, to how light affects the visual pigments in the retina of light), is one of four primary forces. The others are the the eye. Nor is it touched by explanations that regard color strong nuclear force, the weak nuclear force, and as illusory or subjective, a mass hallucination of some sort. gravitation, each theorized to be carried by a different form The hope of unifying relativity theory and quantum of subatomic particle. The strong nuclear force, which theory remains unrealized. But the new understandings prevents the positively charged protons in an atomic nucleus they brought to physics upset classical assumptions, from repelling one another, is carried by mesons; the weak including several integral to the explanation of color. nuclear force, which accounts for radioactive decay and the Absolute space and absolute time fell by the wayside, with emission of beta particles (electrons) by radioactive nuclei, crashes so resounding it was scarcely noticed that the may be carried by a weak boson; the gravitational force, by aether had been discarded too. In part because of the loss a graviton, perhaps. of absolute time, causality became dubious. What was In the proliferation of new theories and new particles, missing was the further reasoning that would have allowed many old questions remain untouched. One is what it means it to be extricated from classical explanations of color, as to say black is an absence, irrespective of whether the in the assertion that light causes color. If no absolute time absence is of light, light waves, electromagnetic waves, exists, an assertion that any A causes any B is meaningless probability waves, quanta, photons, or quarks. Another is or needs further explanation. why light is thought to have an existence apart from color, if Some ideas about time and light were retained-both are one is unvaryingly to be found in the same location as the still said to move-that have never been fully reconciled other. This is not the same question as that of whether color with the explanation of color. If color can be equated with exists apart from light. light, and light moves, why do we regard or perceive color A new question, because the chemical properties of the as unmoving? elements can be traced to the electrons in their atoms, is In the new world of subatomic particles, known why color changes of a chemical nature differ from those primarily through experiments with particle accelerators, that are nonchemical. The statement that, say, orange and distinctions between matter and energy blur. Or the yellow can be mixed to form yelloworange implies that no distinction we have grown accustomed to making is chemical interaction occurs among the substances mixed. In shown to be arbitrary. Like positive and negative space in an example of chemical interaction, copper is orange (or a painting, objects are integrated with the space around copper-colored). Sulfur crystals are yellow. Oxygen is a them. They may be inseparable from it, or accretions of it. colorless gas. Each is a chemical constituent of copper Waves, indistinguishable from particles, are referred sulfate, which for some reason is blue. to as probability waves, and in essence are mathematical We are apparently to conclude that the blueness of abstractions. Photons and fellow wave-particles, in copper sulfate can be explained in terms of the arrangement incessant collision with one another, change their forms of its molecules. Beyond that, why a substance with yellow as they oscillate in time. Fritjof Capra (1975) compared and orange constituents should be blue is a mys this ceaseless activity to a cosmic dance of creation 70 Newton

tery. Color changes in chemical reactions follow intense gravitation from which even light no known rules. And they differ greatly from cannot escape. Whatever the nature of changes observable in nonchemical combining interstellar space, a web of assumptions about of pigments, dyes, lights, or other colored its presumed nullity explains the attention given substances. The neglected question of why this to the problem of how waves of light could pass disparity exists is pressing if both color and through it. I can think of more interesting chemical properties are to be traced to the questions than whether an aether exists. activities of atoms. One is why centuries of debate on invisible Pending solution of the puzzles of the waves versus invisible particles has subatomic world, Newtonian physics, like overwhelmed everyday observation of a Euclidean geometry, is said to provide commonplace visual phenomenon. Ordinary sufficiently close approximations for everyday human beings have taken for granted for use. Most of its basic assumptions are no longer millennia, on the matter of transmission of light, state of the art. But apples still fall out of trees, that it passes through anything transparent oblivious to the news that gravity, no longer a (including vacua), but never through what is force, is now considered a distortion of the opaque. The question of how light, whether curvature of space-time in the presence of wave or particle, is transmitted may reduce less objects. Continuing to explain color in terms of to the question of how media differ from obsolete theory, however, is as misguided as vacuums (or corporeality from incorporeality) navigating a ship by Ptolemaic astronomy, than to that of how transparency differs from though it could probably be done. opacity. An older difficulty, long shunted aside, is The semantics, as often, is revealing. that classical physics never provided a Transparent means transparent to light; opaque sufficiently coherent explanation of color, even means opaque to light. Translucency, an on its own terms. There are too many points at intermediate state, permits passage of light, but which theory cannot be induced to conform not necessarily in a sufficiently coherent form with what we see. The inconsistencies trickle to permit passage of clear images of objects. A down to the largely unsatisfactory formulations recognizable continuum from the entirely of conventional color theory, which relies transparent, through progressively more loosely on classical physics. clouded degrees of translucency to entire opacity, seems correlated with the physical conditions of objects. Transparency, Translucency, Opacity Transparency (ability to transmit light) is a Vacuums, including the interstellar vacuum we functional equivalent for colorlessness, call the sky, have traditionally been regarded whether in vacuums or in material objects. as tracts of emptiness, isolated as if by an invis- Opacity (inability to transmit light) is a func- ible wall from the nonempty portions of the tional equivalent for color. No such thing universe. Interpenetration was inconceivable. occurs as an opaque object that is not, at the Emptiness ceases to be emptiness if it has same time, a colored object. The converse, something in it. Today we know that puta- however, is not the case: color, as in stained- tively empty interstellar space contains a vari- glass windows, is not necessarily opaque. But ety of items. These range from organic its perceived intensity depends on opacity. compounds (about a hundred have been iden- Paintings made with opaque watercolor (gou- tified) to – perhaps -- black holes, regions of ache) look different from those made with Newton 71 transparent watercolor. Ability to reflect light The concepts are interesting to play with (therefore, to appear strongly colored) is a and probably have their own logic. Nothing can direct function of degree of opacity. be both entirely opaque and entirely transparent Vacuums, so far as we know, are always at the same time. The conditions preclude one transparent. And light itself is transparent to another. Anything other than an entirely light, which is why light beams can cross. transparent state is evidently unique to matter, Among inorganic substances, gases are gener- because vacuums are not known to be either ally more transparent than liquids, which are opaque or translucent. We might imagine that more transparent than solids. A transparent matter behaves like a louvered window. The solid, say, glass, is typically opaque when louvers can exclude light entirely, pass it reduced to a powder, a phenomenon explained entirely, or allow it to pass partially. Polarizing in terms of scattering of light by the surfaces of filters, used on cameras, operate in very much the particles. Thickness of layer is usually this manner. A pair of these filters, one significant: a thin sheet of glass might be superimposed on the other so that they can be transparent, a large block of the same type of rotated on a common center, are either opaque glass translucent. Thick or thin, glass can also or transparent depending on how rotated. be opaque, and when opaque is unvaryingly Unlike the filters, most objects are always colored, whether white, black, red, or some opaque, always translucent, or always other color. transparent. Like the colors of objects, their Among organic substances, some change transparency or translucency is presumably to from transparent to opaque under certain con- be explained by the arrangement of their ditions. The change may be irreversible. A molecules. transparent egg white becomes opaque when An absolutely transparent object is abso- cooked. The lens of the eye loses its trans- lutely colorless, displaying only the colors of parency when cataracts form. Honey crystal- the objects behind it. Some opacity is a prior lizes when chilled and becomes opaque in the condition for color. Or color, which was once process. The honey, however, returns to its thought to be an attribute of objects, can also previous translucency when warmed. be explained as a secondary attribute to their The question of what causes light is the opacity, their imperviousness to light. The mirror image of that of what causes continuum of opacity-transparency evidently transparency, translucency, and opacity. In extends beyond the visible light sector of the both color theory and the physical sciences, the . Lead stops X rays, notoriety of the first question has almost although other substances are not opaque to submerged the second. Among a multitude of these rays. Some subatomic particles pass scales devised to measure weight, temperature, through anything. For them, nothing is and other continuum phenomena, I know of opaque. none that enables grading the degrees of The question of whether transparency/ translucency between absolute transparency opacity is inherent in objects or caused by light and absolute opacity. Yet a scale of capability reflects the similar question about color. to transmit light is also a scale between color Because transparency cannot be distinguished (opacity) and its absence (transparency). The from opacity in the darkness, each might be question of what causes opacity is also the called an effect caused by light. More com- question of why light does not interact monly either is regarded as an attribute integral similarly with every material object it meets. to an object, an inconsistent reasoning. 72 Newton

Although a pane of window glass is assumed to The incessant discovering and remain transparent at night, its transparency hypothesizing of particles has been cannot be confirmed under that set of accompanied by a prophesying of qualities for conditions. Like all cats, all panes of glass look these particles through mathematical means. We black (or gray) in the dark. are advised not to inquire what these qualities Would a better explanation of color and light are, other than that they exist or are conjectured evolve if attention were given to the question of to exist and have been assigned interesting tags. why some objects are transparent and other An early example of labeling was regretted. The opaque? We choose our questions for scientific unfortunately named spin of an electron, we are study selectively. On the matter of the often reminded, should not be considered a spin transmission of light, the preference has in any ordinary sense. Unlike Descartes's traditionally been for invisible waves and whirling globules, an electron has no axis on invisible particles rather than the phenomena we which to spin. Spin is now more often see. The impact on the explanation of color has interpreted as a form of symmetry. Some been regrettable. The explanations run around in particles are believed to look the same from all circles. directions, like a steel ball bearing. Others, like In successive reworkings of hypotheses a coin with two faces, have to be flipped twice about its cause, not its nature, color has been to present the same face, and there are other defined in physics in terms of light and light in variations. terms of waves. The waves were identified as To avoid more misunderstandings of the an electromagnetic phenomenon, which Planck kind caused by the electron's spin, and as a and Einstein discovered to be quantized: to replacement for Latin and Greek, whimsical or consist of energy emitted in discrete packets nonsensical names have become the fashion. rather than as an inchoate stream. Between Charm is the tag bestowed on a predicted 1905 and the present, successive peeks behind quality of some or all quarks, which are thought the quantum to find the basic building blocks of to come in at least six flavors. In addition to the universe revealed the photon, fermion, charm, the quark flavors are up, down, strange, boson, hadron, lepton, gluon, pion, meson, bottom, and top (Hawking 1988, 65). Each positron, muon, and other denizens of what has flavor is available in three colors, named as red, been dubbed the zoo of hundreds of subatomic yellow, and blue or red, green, and blue. The particles. These particles may be reducible to jokiness would have left Bertrand Russell quarks, a still lower sublevel of particle that uneasy. The use of color names for predicted also comes in many varieties. The fascinating but unknown qualities is meant to imply, in this ferment has not yet yielded the final word on case, that red, yellow, green, and blue are no what lies behind quarks. We should not await more seriously likely than charm to be the answer with high expectation that it will fundamental characteristics of the physical reveal anything previously unnoticed about world. The real joke might be that no firm color. foundation exists for that assumption.

CHAPTER 9 The Cause of Color and Light

Suppose I say, for instance, "By red light I mean light of such-and- such a range of wave lengths." In that case the statement that light of such wave lengths makes me see red is a tautology, and until the nineteenth century people were uttering meaningless noises when they said that blood is red, because nothing was known of the correlation of wave lengths with sensations of color. This is absurd. It is obvious that "red" has a meaning independent of physics, and that this meaning is relevant in collecting data for the physical theory of colors, just as the pre-scientific meaning of "hot" is relevant in establishing the physical theory of heat. Bertrand Russell, Human Knowledge, Its Scope and Limits

lthough color is said to be caused by eye in one sense but not in another. Because this light, an even more frequent answer is just a vernacular way of acknowledging the to the proverbial question connects paradox, it contributes nothing towards its that idea with another considered equally resolution. Color, like any phenomenon, is selfevident. Color is caused by light and is seen. singular. It either exists or does not, which is In an unwieldly concatenation of the two ideas, incompatible with existing in multiple senses. A color is "the evaluation by the visual sense of simpler understanding is that the two that quality of light (reflected or transmitted by propositions contradict one another. At least a substance) which is basically determined by one is false, incomplete, incoherently stated, or its spectral composition" (American College syntactically meaningless. Dictionary, 1957). The difficulty is more likely to lie with the If color is something that is seen, it requires familiar assertion that light causes color. It the eye. If it is an effect caused by light, inde- leads to too many inconsistencies, whether it pendence from the eye is suggested, a capa- refers to ordinary daylight or light as its veils bility for existing alone. It cannot be both true are progressively lifted through the ever- and not true that the seeing eye is a necessity changing definitions of contemporary physics. for color. To smooth over the inconsistency, The origin of the idea is prescientific. Human color might be regarded as dependent on the beings said color depended on light, rather

73

74 The Cause of Color and Light

than on a functioning eye, long before classi- In explanations of color, these assumptions cal physics evolved with its idealized concep- support the perennial insistence that what we tion of objective light waves rippling through see is not to be trusted and what cannot be seen an objective world contemplated by an objec- is real. Medieval artists painting pictures tive mind. of angels may have believed that angels Our ancestors also had alternate theories. existed. Early theorists on light, color, and nat- As Robert Boyle's seventeenth-century com- ural philosophy undoubtedly believed the pilation shows, they thought color might be universe is full of invisible waves, particles, and caused by invisible waves or equally invisible miasmal. We smile at the innocence of these flying corpuscles. It might be explained as an ancient imaginings, as at the ideas of the interplay of undefined forces (say, a mingling alchemists. But why does modern theory sound of opposites) or the machinations of vaguely similar to ancient theory? The same old lightlike effluvia. Today these ancient schemes invisible waves and particles reappear in more impress as loose language framing wildly con- ornate dress. jectural hypotheses. They also sound uncom- One explanation for the similitude is that the fortably similar to state-of-the-art explanations pre-Socratic Greeks and other early thinkers in the contemporary physical sciences. Defining had an intuitive understanding of the nature of black as an absence of electromagnetic the physical world, which has been elaborated manifestations in the wavelength range of the on in succeeding centuries. They knew nothing visible light sector-or worse, as an absence of about electrons, mesons, gluons, and pions. But probability waves, photons, or quarkscannot they correctly sensed that invisible waves and address the more dubious aspects of defining particles were there, which is exactly what they anything as an absence of anything. What is told us. absent when black is seen is immaterial. We This suggests that essential truth arises want to know what is present. spontaneously from within human The classical question of whether color is in consciousness. What rises with it are ideas that eyes, objects, or light is probably simplistic, as are not true, as in the case of phlogiston and the are monolithic conceptions of causality. If a aether. So we are obliged to separate ideas that single cause must be identified, the eye may are useful from those that are not. I wonder if well have an edge. Light cannot be the cause we can explain what we see without inventing for color, because people with ordinary vision what cannot be seen. do not cease seeing in its absence-we see the The notion that truth lies beyond human darkness. More important, light cannot be the limits is a staple of philosophy and theology. sole cause. Unless the need for a functioning It takes issue with the more modest assump- eye is admitted, we cannot explain why the tion that the only truth available to human blind are unable to see. beings lies in the human condition. Perceptual The model of an objective world is familiar experience, because individual, impresses us from classical physics. It rests on three as a subjectively tainted ephemera, to be assumptions, each considerably tempered in excluded from the objective world of classi- relativity theory and quantum theory. The first cal physics. Russell argued that the exclusion is that objectivity is within the reach of human is "a peculiarity: Physics never mentions beings. The second is that we transcend percepts except when it speaks of empirical subjectivity by deciding to do so, by an act of verification of its laws; but if its laws are not faith. The third is that the gesture is beneficial concerned with percepts, how can percepts because it leads to a higher truth. verify them?" (Russell 1948, 20). The question The Cause of Color and Light 75 is rhetorical; the answer is that they cannot. It is not at all self-evident, from first Explaining red as light with a wavelength principles, that the objective light rays can of 650 millimicrons is incomplete. It includes be completely separated from the sight no indication of how to locate or identify the sense, and that such a fundamental red being explained. A complete statement separation involves very difficult thinking ought to say that the red that we see (the only cannot better be proved than by the way red can be encountered by human beings) following fact. Johann Wolfgang von is accounted for by a particular wavelength. Goethe was gifted with a very scientific That statement, however, is inconsistent for mind (though little inclined to consider the context. A model of the world that pur- analytical methods), and would never see ports to be objective is devoid of any sentient a detail without considering the whole, creature able to see anything. By excluding yet he definitely refused, a hundred years human sensations, an objective world excludes ago, to recognize this difference. Indeed, human beings, because no way exists to sepa- what assertion could give a greater rate one from the other. Within this unpopu- impression of certainty to the lated realm, the phrase "that we see" is unprejudiced than to say that light without meaningless, as is any terminology that simi- the perceptive organ is inconceivable? larly acknowledges sensual experience. Thus, But, the meaning of the word light in this "Physics, per se, has nothing to say about sen- connection, to give it an interpretation sations; and if it uses the word `color' (which that is unassailable, is quite different from it need not do), it will wish to define it in a the light ray of the physicist. way which is logically independent of sensa- Though the name has been retained for tion" (Russell 1948, 261). simplicity, the physical theory of light or A lengthy passage by Max Planck addresses optics, in its most general sense, has as the metaphysical problem of reconciling little to do with the eye and light inconsistencies in the explanation of color and perceptions as the theory of the pendulum light. Light, as the term is used in physics, is has to do with sound perception. This to be understood to possess qualia different ignoring of the sense-perceptions, this from those of light as an experiential restricting to objective real phenomena, phenomenon. which doubtless, from the point of view of immediate interest, means a The first problem of physical optics, the considerable sacrifice made to pure condition necessary for the possibility of knowledge, has prepared a way for a a true physical theory of light, is the anal- great extension of the theory. This theory ysis of all the complex phenomena con- has surpassed all expectations, and nected with light, into objective and yielded important results for the practical subjective parts. The first deals with needs of mankind (Planck n.d., 139). those phenomena which are outside, and independent of, the organ of sight, the For Planck, the two forms of light are com- eye. It is the so-called light rays which parable to apples and oranges. One can be seen constitute the domain of physical and the other is inaccessible to perception. But research. The second part embraces the this amounts to burying the question in words. If inner phenomena, from eye to brain, and light, as the term is understood in physics, this leads us into the realms of physiol- cannot be equated with the light of day, ogy and psychology. identifying one as an explanation for the other is 76 The Cause of Color and Light

a meaningless syntactical exercise. It embroi- to assume that causality can be a meaningful ders on the assertion that color (and, in concept in the explanation of color yet dubious Planck's reasoning, light) is caused by some- elsewhere in the physical sciences. The double thing other than itself. standard is untenable. If time is relative and if If the statement "color is caused by light" is subatomic particles move backward and true in any conventional sense, black should forward in time, we cannot assume a linear not be identifiable as a color not caused by hierarchy in which the phenomena we identify light. Apart from whether true, "color is caused as causes lead to (or antecede) those we by light" may not be syntactically meaningful. identify as effects. The statement asks us to assume something We do not need the virtually incomprehen- called light that, unlike other entities, can be sible computations of contemporary physics to separated from its own color. As Russell's suggest that causality is a concept with only argument implies, that this separation can be limited usefulness, whether in the assertion accomplished is unlikely, however convenient "light causes color" or anywhere else.' The the assumption in some cases. four-year-old child who asks "why?" and who Those of my generation encountered responds to any answer provided with yet Planck's quaint nineteenth-century objective another "why?" is discovering the nature of the world as a rite of passage, a revelation of infinite regress. Reasons, like horizons, can elementary school science class. The objective never be reached. They continually recede as world was a science fiction ghost world, from we approach closer to where they had previ- which human beings had been banished. Spirits ously seemed to be. Even the prime composed of pure thought read instru- antecedent-the idea of God, fate, or big bang as ments with cool detachment and never made the reason behind everything-is not impervious mistakes-or the instruments were studied by to "why?" other instruments. The wonders of this world In the visible light sector of the electromag- included light that could not be seen because it netic spectrum, wavelength is said to correlate had been separated from the dross of sub- with hue, ranging from approximately 780 mil- jective experience. But how can we find objec- limicrons (red) to 380 millimicrons (violet). tive light if it cannot be seen, if it exists This correlation is unlike that of, say, the ther- "outside, and independent of, the organs of mometer, which measures a smooth transition sight?" from extreme cold to extreme heat. The sectors Asserting that light exists in an unseen of the electromagnetic spectrum are unlike one objective form shores up its respectability as a another. Ultraviolet rays pass into X rays, causal explanation of color. The visible is which have different physical characteristics. shown to have invisible forces behind it, and No intermediate range of rays has some of the the invisible is what counts. If asked what characteristics of X rays combined with some causes a green color patch I see, I am not of the characteristics of ultraviolet rays. expected to answer "green paint." The correct In the visible light sector, the colors of the answer is objective light, the Planckian light I electromagnetic spectrum are those of the rain- cannot see, although I should know it is there. bow, arranged in the same order. Wavelength If it were not there, how could I answer the does not correlate with those visual attributes question "what causes color?" of hue that can be identified. The brightest part These curious assumptions about causality of the spectrum is yellow, near the center. have a further dimension today. We are asked Yellow-green, also near the center, is said to The Cause of Color and Light 77 be the range in which the finest hue would be considered an absurd definition, discriminations can be made. The pair of because inappropriate or too general. Yet, quite complementary hues with the smallest regularly, colors are said to be wavelengths of difference in color value is red and green. The light, probability waves, quanta, photons, pair with the greatest is yellow and purple. bosons, quarks, or similar subatomic units. Our None of these visual effects can be genuinely ability to describe elephants vastly exceeds our correlated with wavelength. ability to describe colors. Russell's objection to the exclusion of The exclusion of percepts from physics is a percepts reaches to the heart of one difficulty secondary phenomenon. Physics purports to with explanations of color that rely on classical examine reality, however that might be defined. physics. Imagine that a scientist from another Percepts, by unspoken agreement, are usually galaxy knew that certain wavelengths of light excluded from our conception of an objective were supposed to look green. The scientist was reality. For this reason, the question of whether unable to verify that this was correct because he color exists in light, in the eye, in objects, in or she had no idea of the meaning of the word some or all of these, or (as Plato imagined) in green or of the nature of the difference none of them superficially resembles the bogus supposed to exist between green and red. How enigma of whether the sound of the crash exists could this visitor be edified about what red and when a tree falls in the forest. The intended green look like if no samples were available to question is whether human consciousness can show? be the criterion for reality or for those realities The task evidently cannot be accomplished: accessible to human beings. It is culturally color is apparently so innate to the sensory biased in that only the unheard sound is realm that words are insufficient to convey its assumed to be conjectural. The thought nature to those who have never seen it. Yet experiment of Schrodinger's cat raises the people regularly learn about items that they question in quantum theory: We are asked will never see. And dictionaries exist because whether a cat, unseen by any observer able to word definitions are not only possible but verify its condition, can be regarded as alive or adequate to the purpose in most cases. dead for any observer other than itself. We are Dictionaries are able to describe an elephant not invited to consider whether (or why) the as a large quadruped with tusks and a trunk. cat-or the tree, the forest, or the event of the Yet they are limited to defining colors either by fall-should be assumed to really exist in the simile or in terms of light waves, a conceit that absence of a verifying observer. Events and is a simile of another variety. "Blue," for objects are assumed to be real, but perceptions example, has been called "a color whose hue is of them may not be. Alfred North Whitehead [like] that of the clear sky or [ like ] that of the reported that Galileo, to whom the puzzle of the portion of the spectrum lying between green tree was put in a more complicated manner, and violet" (Webster's Third New International "considered this question, and at once pointed Dictionary). Dictionaries are not reduced to out that, apart from eyes, ears, or noses, there defining elephant by telling what the beasts would be no colours, sounds, or smells" resemble or where some of them can be (Whitehead [1925] 1953, 55). located. Russell would have approved of the Nor are we asked to accept, as sufficient to answer, which is consistent with the Copen- the question, that an elephant is an aggregate hagen interpretation of quantum theory. of molecules, as is a cat or a mouse. That Unlike classical physics, this holds that the 78 The Cause of Color and Light

observer is part of any experiment and that (if the fall occurred). Or it can mean the effect reality cannot exist in the absence of a verify- of these waves on a listening ear or recording ing observer. A reality that depends on the device. observer comes close to implying that the only The question of whether color is in the eye, reality is that of the observer. The dilemma in light, or elsewhere is only partially analo- might be resolved by assuming the possibility gous to that of the subjectivity or objectivity of of multiple observers. We can also ask what any sound caused by the fall of the tree. The reality is intended to imply. explanation of sound as a wave phenomenon is A simpler approach to the question about considerably more coherent than the the falling tree is familiar to modern school- approximately parallel reading of color as a children. The answer depends on how sound wave of another sort. This can be seen by turn- is defined. The word can mean the sound ing to the particulars of how light is held to waves assumed to have been caused by the fall cause red-violet, brown, blue, and black. CHAPTER 10 Red-Violet, Blue, Brown, and Optical Mixture

If red and violet be mingled, there will be generated according to their various Proportions various Purples, such as are not like in appearance to the Colour of any homogeneal Light, and of these Purples mix'd with yellow and blue may be made other new Colours. Sir Isaac Newton, Opticks Purples and magentas do not exist, they are a figment of our perception. Anton Wilson, "Film Feedback'

hile other societies have developed icists, like everyone else, identify the spectral different methods of naming, the hues as red, orange, yellow, green, blue, and major colors of the rainbow or violet. They agree about the continuum nature solar spectrum are usually identified today as of the range and that a diagram can be red, orange, yellow, green, blue, and violet. constructed. Newton included indigo, a seventh hue intermediary between blue and violet. The major spectral colors, sometimes called major Diagrams of Hue Relationships hues, exist in a continuum relationship. What kind of diagram best shows the Between, say, blue and green a range of relationship among the hues? If color is intermediaries is found. These are called bluish regarded as an effect of light waves, one kind of greens or greenish blues, according to whether diagram is used. If classified as a percept, they more nearly resemble green or blue. The another kind of diagram is required. The two relationships among the major hues are diagrams are incompatible. If the nature of the familiarly shown in diagrammatic form. relationship among the hues is correctly shown None of these commonplace concepts is by one, it cannot be correctly shown by the incompatible with scientific theory, whether other because one is linear and the other circular in its classical or more recondite forms. Phys- (figure 10-1).

79 80 Red-Violet, Blue, Brown, and Optical Mixture

INFRA-RED

RED ORANGE YELLOW GREEN BLUE VIOLET

ULTRA-VIOLET

Figure 10-1. The electromagnetic scale and the color wheel. The electromagnetic wave scale and the color wheel are topologically inconsistent because one is linear and the other is circular.

The diagram familiar from classical physics green, and blue, between red and violet, appear is unmodified by more recent developments. in the visible light sector in the order in which Following Maxwell's reasoning, it assumes that they can be seen in the rainbow. In variation in color among the spectral hues is electromagnetic theory, the color or visible light correlated with variation in wavelength of sector is part of a "longer" scale, prop- light. The range of wavelengths accounting for erly diagrammed as a line segment for this visible light (color) is only a short segment of reason. the electromagnetic scale. The scale also includes gamma rays, X rays, ultraviolet, infrared, radar, radio and television waves, Color as Percept microwaves, and any other form of radiant When color is regarded as something we see, a energy. percept or visual phenomenon, a different The visible light portion of the scale consists diagram is used. This second diagram is the of wave lengths from roughly 380 to 780 familiar color wheel or color circle. Newton, millimicrons (3>800-7,800 angstrom units). The credited with its invention, "became surprised to lower extreme consists of violet light (380 see [the spectral colors emerge from the prism] millimicrons); the upper extreme, of red light in an oblong form, which, according to the (780 millimicrons). Beyond red lies infrared. received laws of Refraction, I expected should Beyond violet lies ultraviolet. Infrared and have been circular" (Phil. Trans, 6:3075). On ultraviolet are said to be invisible to human the color wheel, red and violet, no longer beings, though Helmholtz reports that some representing extremes, lie adjacent to one people can see ultraviolet. Orange, yellow, another.

Red-Violet, Blue, Brown, and Optical Mixture 81

The compelling argument for the color The Second Continuum wheel's arrangement is experiential. We see a In diagramming the relationships among the continuum of colors intermediary between red hues, the locus of the topological incongruity is and violet. These colors have such names as the relationship between red and violet. These reddish violet and violet-red, though Newton colors either should or should not be preferred to call them purples. Infrared, envisioned as extremes, as opposite ends of a ultraviolet, and the wave phenomena lying linear scale. In topological terms, lines and cir- beyond each are not included in the color cles are related. A line segment can be created wheel. The relationship among the hues is by cutting the perimeter of a circle, just as a regarded as circular. circle can be created by joining the ends of a Red-violet is left in an ambiguous position line segment (figure 10-2). by the disparity between the color wheel and Imagine the color wheel severed at that the electromagnetic wave scale. Although a point on its perimeter where red and violet hue variation recognized on the color wheel, it meet. The question is whether one or two color has no place on the electromagnetic spectrum continua lie between red and violet. Are two or in scientific theory. No way exists to add different cities (call them red and violet) joined red-violet to the scale of wavelengths. Red- by two roads or just one? violet belongs at the top of the electromagnetic If color is to be equated with wavelength (if scale and also at the bottom. Yet it cannot be the relationship between the spectral hues is added in both places without violating the lin- linear), a single continuum or road is implied. ear nature of the scale. In scientific theory, red It consists of orange, yellow, green, and blue, grades into infrared, and violet grades into those hues assigned wavelengths between the ultraviolet. As we see color, red-violet provides a transition between red and violet.

Circular and Linear Diagrams The color wheel and the electromagnetic scale suggest that the entity called color can be sub- divided into half a dozen major hues. The rela- tionship among the hues is analagous to the perimeter of a circle (it can be diagrammed in circular form). But it is also analagous to a line (it can be diagrammed as a line segment). This is a logical inconsistency and a topological incongruency. Because circular means non- linear (and linear mean noncircular), no set of relationships can be both circular and linear at the same time. If, say, biological evolution is regarded as a linear (hierarchical) transition Figure 10-2. Transforming a circular from lower to higher life forms, we cannot diagram into a linear diagram. A cir- consistently assert that its nature is also circu- cle can be transformed into a line by lar (cyclic): that higher forms, say, mammals, cutting it-or a line segment cannot both meet itself end to end to form a will eventually develop into forms lower on circle and not meet itself at the same the evolutionary scale, say, protozoa. time.

82 Red-Violet, Blue, Brown, and Optical Mixture

extremes of red and violet. What we see tells difference must be identified, if any, between a us, however, that two roads exist, recognized in color that exists-yellow-and another seen the circular arrangement of the hues on the alongside it that does not-red-violet. Confu- color wheel. The continuum additional to sions proliferate if Newton's experiment is orange, yellow, green, and blue consists of repeated with tubes of red and violet paint. those colors commonly said to lie between red Red and violet both exist and are each part of and violet: the range of red-violet colors. the electromagnetic spectrum. Yet mixing the Which diagram provides a more accurate two colors of paint produces red-violet, a model for color relationships? I conclude the "nonexistent" color. Adding to the red-violet color wheel is correct and the diagram of paint small quantities of other existing wavelengths is incomplete. By failing to colors -- yellow, green, brown – produces include red-violet colors, it implies they do not mixed colors that partly exist and partly do or cannot exist. This is incorrect because we not. see a large range of red-violet colors. How are What meaning can we extract from these we able to see them if they are not on the bizarre propositions about the nature of red- scale? violet? If color itself is illusory, what can it Conjecture runs rampant on this question. mean to classify red-violet as more illusory Red waves and violet waves, we are told, prob- than, say, orange or green? No matter how ably mix in the eye in some manner. This mix- rearranged, the argument that shades of color in ing enables us to see red-violet colors or to the red-violet range are "figments of percep- imagine we see them. Unlike other hue varia- tion" falsifies the theory it purports to sustain. tions, they cannot be correlated with any range All colors or all hue variations cannot be of wavelengths of light. To buttress these sup- explained in terms of wavelengths of light if positions, we are asked to make a curious the range of red-violet colors requires a differ- assumption: the red-violet colors are not really ent explanation. colors, or do not really exist (as colors). Thus, Putting forward optical mixture as the alter- theory continues, we can salvage the idea all nate explanation for red-violet colors is tanta- colors or variations of hue can be correlated mount to an admission that the color wheel is with wavelengths of light. Red-violets do not correct and the diagram of wavelengths is fit in the picture because they are not really incomplete. Two color continua lie between there. red and violet in the experiential world. The The explanation, now woven into folklore, ideal world of classical physics can include is too slapdash to take seriously. The thesis of neither red-violet nor variations on colors in optical mixing in this sense is untenable. There that range: it is assumed to include no human is no such thing as colors we see that are not eye in which red and violet waves might mix. hallucinatory but nevertheless do not exist. No way can be found to give a reasonable expla- nation of what such a statement could mean. Circular and Linear Relationships Because color is a visual phenomenon, the In electromagnetic theory, which became a criterion for whether a color exists is whether cornerstone of classical physics, red-violet was that color can be seen.' sacrificed to preserve an idea that cannot be Any attempt to dispose of the initial diffi- reconciled with visual observation. The rela- culty leads to others that are worse. Consider tionship among the hues, as Newton recog- the hypothesis that we actually see yellow but nized, is circular rather than linear. Theory only imagine red-violet. To define this idea, the does not match observation, and the degree Red-Violet, Blue, Brown, and Optical Mixture 83

of disparity is unusual. I can think of no other can be explained in terms of polarities, say, instance in which the quality to be measured more-less, bigger-smaller, longer-shorter, or by a linear scale appears not to be linear, there- better-worse. Each end of the line segment is fore, to be incapable of measurement by that labeled to correspond to one of the extremes. kind of scale. For the diagram to be useful, a correlation The quality experienced as heat is measured should exist with an imaginable, even if never by the thermometer. The thermometer, which experienced, extreme in perceptual exper- is linear, shows hot and cold as polarities, with ience. a single continuum between these extremes. The heat, pitch, and weight scales can be The linearity of the thermometer and its single correlated in this manner. Therefore, no con- continuum are compatible with perception. We ceptual barrier exists to envisioning the exten- cannot imagine, and have never experienced, sion of each. I can imagine more heat than I any condition intermediary between hot and have ever felt, higher pitched sounds than I cold except for a transitional warmness. There have ever heard, weights heavier than I can lift. is no other intermediary continuum that is The wavelength scale, unlike these other dissimilar to warmness. scales, cannot be correlated with extremes in Many further examples could be listed. the perceptual experience of color. Imagining Pitch, which the ear perceives in sounds, is the linear extension of the scale is impossible. envisioned as linear. Sounds range from low Red is not perceived as more color than violet, pitched to high pitched, and perception again though its wavelength is longer. Nor does red agrees with theory. Extremes of high-pitched look longer or brighter-it cannot be sound do not grade into low-pitched sound as characterized by a superlative of any quality they grow progressively higher. found to a lesser degree in violet. Red just Weight ranges from light to heavy, a linear looks different from violet, and either looks conception. Objects that are becoming heavier different from yellow. do not, at a certain point, circle back to If human beings were able to see by become lighter. Heat, pitch, and weight are infrared light, which has a longer wavelength each properly diagrammed by a line segment, than red, we cannot say that this experience not a circle. In none of these instances is the- would make available more color or more ory inconsonant with perception on the ques- extreme color than we are able to see now. I tion of whether linearity or circularity can only imagine that infrared might look characterizes the set of relationships. different from the colors we presently see, just The linear nature of the electromagnetic as each of these colors looks different from any scale introduces more than just ideas about the other color. hues that are irreconcilable with visual The high and low extremes of the visible experience. The scale implies conditions light sector of the electromagnetic wavelength about color that are impossible to imagine scale cannot be correlated with extremes of because they make no sense topologically. A visual experience. Red is not an opposite to line segment, for example, has two ends, violet, in the sense that hot might be thought although the perimeter of a circle has none. an opposite of cold, high pitched of low Because of this two-endedness, line segments pitched, or heavy of light. Nor is color linear are used to construct diagrams that serve as or binary. We find color, in electromagnetic visual analogues for hierarchical relationships. theory, forced to a pattern it does not fit, the In these relationships, a range exists exact reason the range of red-violet colors between dissimilar extremes. The extremes appears to have been mislaid.

84 Red-Violet, Blue, Brown, and Optical Mixture

For Newton, "the received laws of produced which is red-violet in color is not Refraction" suggested a circular relationship clear. among the hues. But Maxwell, in devising the electromagnetic scale, was interested in another Seeing Ultraviolet set of relationships. He hoped to show that The status of red-violet, in electromagnetic light, radio waves, X rays, and other theory, is no more curious than that of phenomena could all be explained as forms of ultraviolet and blue. Wavelengths in the radiant energy of varying wavelength. ultraviolet range, like those in the infrared Red-violet fell victim to the synthesis, and the range, are said to be imperceptible to human electromagnetic scale, for this reason, falls short beings. The lens of the human eye filters them in its explanation of the hues. out, a mechanism suggesting that ultraviolet Can we retain the hue circle and discard the rays are damaging and the retina is sensitive to electromagnetic scale? The hue circle has no them. room for infrared, ultraviolet, and other forms Despite the filtering (which appears to be of radiant energy beyond the visible light partial), experimental evidence is available that sector. In theory, the electromagnetic scale human beings, and not just some insects, are might be adjusted. No barrier exists to a line able to see by ultraviolet light. The literature is segment on which two points are joined by two divided only on the question of whether the lines (figure 10-3). Negative numbers, from capability is commonplace. Some writers said -380 to - 780, could be used for the red-violet that it is (Jenkins and White 1957, 202). Others sector. But this introduces the conception of contended the ability is limited to persons who negative wave lengths, which is untenable. And have had the natural lens of the eye removed, as whether a monochromatic beam can be in cataract operations. Irrespective of the size of the human population involved, those able to see by ultraviolet light do not perceive its color as violet, purple, lavender, or any of the other shades misleadingly suggested by its name. The color is described as bluish gray. Helmholtz filled this out by observing that when ultraviolet rays are of low intensity, "their color is indigo-blue, and with higher intensity bluish gray" (Helmholtz [1909] 1962, 2:66). The color of ultraviolet rays has a bearing on the location of blue. Although red-violet has no place on the electromagnetic spectrum, the

Figure 10-3. Alternate construction for the visible blue range, which should be found at one place, light sector of the electromagnetic wave scale. Two occupies two. The perceptual experience of points on the electromagnetic wave scale, A and B, are placed at 380 and 780 millimicrons, the extremes seeing bluish gray (or indigo-blue) may of the color, or visible light, sector. A and B are indicate that waves from the blue sector of the assumed to have two discrete circuits, or paths, scale have entered the viewer's eye. It can also between them. This arrangement allows room not only for the spectral colors presently on the scale but indicate, relying on Helmholtz, that the viewer also for the red-violet range, which is presently is seeing ultraviolet, waves from a different excluded. range that is not adjacent to blue. Why should

Red-Violet, Blue, Brown, and Optical Mixture 85 they resemble blue more than they resemble see, if they see anything, when encountering violet? wavelengths in the ultraviolet range. That waves from more than one range look blue suggests a disruption of what is usually The Cause of Brown regarded as a one-to-one relationship between Newton differentiates between "all the Colours the visible hue continuum and the electrom- in the Universe which are made by Light" and agnetic continuum. But phenomenological real- those that depend "on the Powers of ity is more complex than theory, and whether Imagination . . . as when by the power of all sectors of the electromagnetic spectrum exist Phantasy we see Colours in a Dream, or a in a smooth linear relationship is not cer- Madman sees things before him which are not tain. The red extreme of the visible light sec- there; or when we see Fire by striking the Eye, tor passes into infrared, radar, and short radio or see Colours like the Eye of a Peacock's waves. In 1917, Nichols and Tear produced Feather, by pressing our Eyes in either corner infrared with wavelengths up to 0.42 mil- whilst we look the other way" (Newton [ 1730] limeters; radio waves down to 0.22 millimeters 1952, 161). The colors made by light include ( Jenkins and White 1957, 203). those of natural bodies, and "every Body That this could be accomplished suggests an reflects the Rays of its own Colour more overlap between these two sectors. Wave copiously than the rest" (Newton [1730] 1952, phenomena in the zone of overlap (0.42-0.22 179). Newton demonstrated this by his millimeters) can be either radio waves or seventeenth experiment, which tests the infrared. Or wavelengths that in theory ought to reflective powers of vermilion (cinnabar) and have been infrared were shown to be radio ultramarine pigment under red and blue lights. waves and vice versa. The apparent disruption Each pigment looked brighter under the in the far infrared sector of the scale suggests light that resembled it in color. It looked darker parallels with the apparent bifurcation of blue or duller under the other light. Newton's (or indigo-blue) at the other extreme. observation that the experiment works best The earth's atmosphere filters out ultraviolet with objects of "the fullest and most vivid rays, as does the lens of the human eye. This Colours" (Newton [ 1730] 1952, 180) ought to filtering suggests that the rays, linked to the have warned of the danger of generalizing to formation of skin cancers, are damaging to the "all the Colours in the Universe" other than retina in large amounts. In that case, a visual those thought to be phantasmagorical. pigment adapted to responding to ultraviolet Vermilion, a red pigment, reflects pri- rays would be superfluous, though it might marily red light. Ultramarine, a blue pigment, occur as an evolutionary relic. Ultraviolet rays reflects primarily blue light. But neither pig- entering the human eye must interact with the ment reflects light of a single color or single mechanism for seeing blue or for seeing blue wavelength. The deviation is the basis for and violet. Helmholtz describes the visual modern spectrophotometry and colorimetry, sensation as indigo-blue, and Newton named which assess the multiple colors of light that indigo as a seventh hue between blue and most surfaces reflect. Newton's understanding violet. of his seventeenth experiment implies that My dictionary identifies blue as the color ideal red and blue objects can exist, function- "between green and violet in the spectrum." ing as ideally selective reflectors of red and Blue arguably is also the color lying beyond blue light. But the objects of the world are violet, in the range called ultraviolet. Indigo- imperfect reflectors that fail to conform to the- blue or blue-gray are the colors human beings ory in this form. 86 Red-Violet, Blue, Brown, and Optical Mixture

Multiple reflectance is at the heart of why ficiality of the explanation can be gauged by its Newton's conclusions from his seventeenth vagueness. What mixing in the eye (in what experiment cannot be expanded to explain the part of the eye?) implies is unclear, though it colors of brown or metallic-colored objects. suggests the bare minimum that the rays enter No sense exists in which any brown or metal- the eye simultaneously. Why should a mixing lic object "reflects the Rays of its own Colour." internal to the eye differ from an external Brown and, say, copper are not included in the aggregating? If rays of light mix in the eye to visible light portion of the electromagnetic enable us to see brown, taupe, olive, russet, and spectrum. Brown spectral rays do not exist, similar colors, no barrier should exist to nor do silver, bronze, copper, or gold spec- bundling the same rays outside the eye to tral rays. produce beams of monochromatic light in these In modern colorimetry, the issue of how colors. Yet this cannot be done. brown is to be related to the colors of the spec- The easier conclusion is that the trum has been addressed by reinterpreting it generalizations that can be drawn from as a supersaturated yellow. The reinterpre- Newton's seventeenth experiment are limited. tation is unconvincing on epistemological The colors of some substances, say ultramarine grounds. Supersaturated, a term borrowed and cinnabar, can be explained by assuming from chemistry, has no clear meaning in rela- they reflect rays of their own color. The colors tion to color. In chemistry, it refers to the of most substances, objects, or surfaces cannot behavior of liquids in which substances are be explained in this manner. More than one dissolved. How are we to supersaturate a ray of wavelength of light is reflected, and the light? Furthermore, that all browns look similar electromagnetic spectrum may include no to yellow is not a viable assumption. Many monochrome wavelength matching the color of browns are greenish, reddish, orange, even that object. The large range of colors called purplish. brown is far beyond the scope of Newton's explanation. Brown color in objects cannot be correlated with a component in light that Optical Mixing matches the color of the object and is reflected Browns are usually classified in color theory from the object's surface. not as supersaturated yellows but as tertiary colors, mixtures that include each of the primary colors. No monochromatic wavelength Optical Mixture exists for any brown, just as none exists for Mixing in the eye is put forward too often as red-violet. Instead, the browns, chromatic grays, an explanation for color phenomena it does and other colors that approach brown or neutral not explain. I do not mean to call into ques- are created by mixtures of varying wavelengths. tion, however, what the Impressionist and The explanation wobbles when we are told the neo-Impressionist painters regarded as optical mixture occurs in the eye, that ubiquitous mixture (melange optique). The term, bor- explanation of too much. Eyes are presumed not rowed from Maxwell and Rood, refers to an to be present in the classical model of the observable visual phenomenon. Its foundation physical world. is that no visual analogue can be found for Mixing in the eye has become the default emptiness, just as none can be found for explanation for color, put forward to explain silence. We see color throughout the visual field how we see colors that cannot be correlated even when the limits of visual acuity are with a single spectral wavelength. The super- exceeded. Red-Violet, Blue, Brown, and Optical Mixture 87

Two colors on a spinning Maxwell disc, for else, of the colors of the disc at rest, which are example, can be rotated too fast for either to be assumed to be the real colors of the disc. seen individually. Unlike the rapidly mov- ing As in the Newton-Hooke argument about wings of a hummingbird, the colors on the whether the spectral colors are inherent in light, Maxwell disc do not become invisible. Instead, we need not assume that objects possess normal the observer perceives a third color, classified colors that they display under normal as an optical mixture of the colors spinning on conditions. Because color in objects is the disc. contextual, the dark sky of night is as normal as A similar phenomenon occurs when objects the blue sky of day. The color of the disc when are so small that they pass below the threshold spinning is its real (and only) color for that of conscious perception. A building made of tan condition. bricks and brown bricks may be too far away to Arguing that a distant building, which allow its individual bricks to be seen. The appears to be a single color, is actually building becomes neither invisi- ble nor no multicolored (or made of multicolored bricks) is color, even if the color spots that correspond to as arbitrary as arguing that the building is really its bricks lie below the thresh- old of vision. It made of protons, neutrons, and electrons, which presents a color of its own, again assumed to be human beings would see if we were small an optical mixture of the colors of the individual enough. The lesson of optical mixture is that bricks. color is unitary in nature. The seeing eye always After a century of critiques of Impres- sees colors but never more than one at any sionist and neo-Impressionist painting, the term given time and place. The visual field can be optical mixture is too firmly entrenched in the modeled as a two-dimensional matrix, a grid of English language to be dislodged. But it may boxes resembling a piece of graph paper. We be redundant. To regard the single color might imagine that the boxes are tiny, each box presented by a spinning Maxwell disc as a mix- contains a color, and no box is able to contain ture implies it must be a mixture of something two colors at the same time.

CHAPTER 11 Achromatic Colors and Mirrors

Black is a real sensation, even if it is produced by entire absence of light. The sensation of black is distinctly different from the lack of all sensation. A spot in the field of view which sends no light to the eye looks black; but no light comes to the eye from objects that are behind it, whether they are dark or bright, and yet these objects do not look black, there is simply no sensation so far as they are concerned. Erwin von Helmholtz, Treatise on Physiological Optics

ntimations of the negative and backward quently cannot be a color. I submit in place of characterize most statements about black, this confused reasoning that, because black is a the retrograde color which is variously color, light waves do not cause all colors. The nothing, not caused by light, or not caused by more usual or more popular train of thought the phenomena that account for light. Like forces an untenable conclusion. It implies that brown, silver, or red-violet, black confutes black is a noncolor produced when light acts as Newton's conclusions from his seventeenth a noncause, a conceptual morass in which experiment. No sense exists in which a black negative events result in negative consequences. object, or any object that is achromatic in color, If causality is to be assumed and black is not reflects "Rays of its own Colours." To smooth caused by light, a categorical hiatus is created out these discrepancies, further negatives flesh unless its formation can be attributed to some out the picture. other agency. As both Helmholtz and the Buddha have wisely observed, seeing black is not tanta- The Cause of Black mount to not seeing. Black is a color, as is The popular assertion that black is not a color white or gray, by any reasonable visual rests on the syllogism that light waves cause criterion. In common with all other colors, colors, yet do not cause black, which cones- black is exclusive to the visual field. We see

88

Achromatic Colors and Mirrors 89 it and do not hear it. Like any color, black form/space. Although great care is needed in meets the criterion of exclusivity in spatiotem- distinguishing between two-dimensional and poral location: I cannot see black at the same three-dimensional space, color and space are time and place I see yellow. both devoid of corporeality but exhibit dimen- The turgidity of popular lines of reasoning sion or extension. The valence of form is posi- about black conveys its own message. We tive, perhaps because Plato idealized the forms must look to culture before science to under- of geometry. The valence of color is negative. stand the conception of this color. Salvaging In the familiar web of symbolic associations, the presumed oppositeness of black and white, form is "real." It is structured, analyzable, an idea originally inspired by primitive recog- intellectual (perhaps because associated with nition of the difference between night and day, geometry), rational, and consequently is integral to our conception of oppositeness masculine. Color (or color qua space) is in general. The salvage job forces us to sustain chimerical, unreal, unstructured, unanalyzable a conception of black inconsistent with visual (hence, suggestive of chaos), emotional, experience. irrational, and consequently feminine. Recognizing that black is a color-because Over the centuries, layers of additional we see it as a color-raises questions about the association have been piled atop these. Angu- generalization that colors can be correlated lar forms are said to be masculine, curved with wavelengths, thence questions about forms feminine, as if gender should be whether electromagnetic theory provides an attributed to geometrical shapes. Among non- adequate explanation of color. Recognizing corporeal entities, light is regarded as black as a color-assessing it according to vis- immaterial but real. Colors are immaterial but ual criteria-casts in doubt the presumed unreal, presumably because they are second- oppositeness of black and white. Oppositeness aries that are classified as effects of light. Hue of color is not a visual concept. is opposed to [color] value, the rationale for When the world is interpreted as a collec- differentiating between chromatic and achro- tion of opposites (black and white are among matic colors. Among the achromatic colors, the pairs of opposites), each pair is imagined black, reduced to no more than an antipode to consist of a positive member poised against for white, is defined almost totally in negatives. a negative twin. In psychological association It is not a color, not caused by light, and it is and in symbolism, positives may be freely interpreted as nothing when seen at night or interchanged, as may negatives. Black, for this in interstellar space. Factors more profound reason, is said to symbolize (or suggest) night, than the foibles of fashion explain why the darkness, the void, Satan, evil. In our society walls of houses and apartments are often black is rarely associated with day, light, sal- painted entirely white but rarely entirely black. vation, God, goodness, or other concepts more Anyone so inclined may write sonnets in intimately linked with white. Gray has few praise of colors, including black. But the nega- symbolic associations, an omission I consider tive member of each opposed pair is subtly significant. A world seen predominantly in devalued, flagged as less worthy of respect or terms of black or white, true or false, allows of serious attention. Oppositeness is a value little room for the continuum of ambiquities, system. And color, primarily because other for that which is more or less true, or neither than form (or because of its association with exactly black or white. space), drifts to the negative side of the bal- Among the traditional pairs of opposites, ance. Even if acknowledged to be good to look form/color is virtually indistinguishable from at, it may be dismissed as just decorative, as

90 Achromatic Colors and Mirrors

in early critiques of the use of color in the Figure11-1.The spectral colors and their wavelengths according to James Clerk Maxwell. Note that Maxwell' paintings of Henri Matisse. wavelength units are inch units rather than millimicron Within these stereotyped limits, color is not and should be multiplied by 10-8. (After Maxwei often expected to be of significant interest to [1890] 1965, 425). serious thinkers, or even to manly men. Spectral Color Wavelength Although the merits of draftsmanly (form- oriented) versus painterly (color-oriented) red 2,450 painting styles have historically been consid- scarlet 2,328 ered at greater length than so trivial a topic orange 2,240 deserves, I do not mean to imply these issues yellow 2,154 are often debated. Rarely are they even given yellow-green 2, 078 much thought. But we live out our stereotypes green 2,013 as we sense them, often teetering on the edge of green 1, 951 reducing human potential to a cartoon. bluish green 1,879 Trivial in the instance but telling in the blue-green 1, 846 aggregate, thoughtless comments about color greenish blue 1,797 are made even by those who we assume have blue 1, 755 thought about it more deeply than others. blue 1, 721 Newton discoursed on "the more eminent blue 1, 688 colors" without explaining his criteria for color indigo 1,660 eminence. Maxwell's list of the colors of the indigo 1,630 spectrum and their corresponding wavelengths indigo 1,604 is apocryphal for its omission of violet and its tortured use of common color names (figure 11-1). Bluish green is not usually regarded as a Oppositeness was originally a philosophi- less blue variety of blue-green. Scarlet is cal concept, not a scientific discovery. But it ordinarily ranked as a subclass of red, not, as long ago became incorporated into theory in Maxwell has it, as a major hue name of the level the physical sciences, assuming a position so of red or orange. These small oddities suggest a central it survived even the demise of New- degree of societally determined indifference to ton's absolute space and absolute time. Today, color or to visual experience. parity is regarded as a fundamental feature of The indifference is difficult to comprehend the subatomic world, as if an orderly God had because inconsistent with the human condi- ordained that everything ought to come in tion, with what the mass of human beings pairs. really feel and think. The prospect of blindness In contemporary physics, manifestations of evokes terror. Freud equated loss of the eyes parity include the charges of subatomic parti- with castration, pointing to, say, Oedipus Rex cles (positive, negative, or no charge), which as evidence of the severity of the punishment, cause them to attract or repel one another. fitting for only the gravest sins. For the pro- Each particle is said to be paired with an tagonist of Rudyard Kipling's The Light That antiparticle so opposite to itself that it may Failed, loss of the ability to see meant loss of even insist on moving in an opposite direction a reason for living. We are not genuinely through time. A chance meeting of particle and indifferent to what we see. We care greatly antiparticle results in the annihilation of both. whether we see. What else would opposites do?

Achromatic Colors and Mirrors 91

The photon, which is said to account for coherent. A white surface reflects only light. A light and color, fits uneasily into this system. mirror reflects light reflected from the surfaces The photon is described as without mass, of the objects in front of it, in a manner without charge, and as its own antiparticle. The allowing us to see images of these objects. semantic question is how functioning as one's Mirrors are typically silver, not white. Color, own antiparticle-one's own opposite-differs rather than substance, is critical. from having no antiparticle. I conclude that the Ordinary household mirrors are "silvered" possibility that all particles may not have with mercury. Aluminum compounds coat antiparticles is regarded as threatening, and movie screens, the legendary "silver screen." tortuous explanations are constructed to avoid Although all highly polished surfaces are considering the possibility. Can the photon reflective, the best mirrors are created by annihilate its antiparticle on meeting it, if the objects that are either silver or metallic in color. photon is its own antiparticle? The polished shield used as a mirror is a familiar fixture in myth, as in the Greek tale of the slaying of the Gorgon. The roofs of houses, Absorbers and Reflectors of Light to enable them to reflect light, are given Black and white are sometimes explained in coatings of silver-colored substances, known to terms of an interaction between light and be more effective than white. Can any empirical objects. Thus, the ideal black object is said to reason be found for classifying white, not silver be a perfect absorber of light. The ideal white or a mirror, as the ideal reflector? I doubt it. The object is a perfect reflector. The ideal conventional reasoning is meant to sustain the transparent (colorless) object is a perfect theory that colors have opposites, and that transmitter. The definitions survive because white, not silver, is the opposite of black. they make the behavior of light eminently easy to understand. We all know the diagrams in which waves, usually represented by arrows, Causes of Transparency are shown bouncing from surfaces as if they Although Munsell complained about the were volleyballs. ambiguity of color names, terms that refer to As is generally true of idealized achromatic or transparent conditions are equally conceptions, the parameters are too simplistic ambiguous. Misuse of the terms, or confusing to help us understand visual phenomena. Black usage, is common and institutionalized. Physics cats look black in the daytime, ostensibly from has its study of black bodies. These are not absorbing all light that falls on them, but as necessarily black, raising the question of why gray as everything else in the dark when no they are known as black bodies. The mutilation light is present that they might absorb. A of language might be traceable to Max Planck, blackboard ought to be more efficient absorber whose conception of "the black heat rays of light than a white wall. But either surface emitted from a stove" attributed color (which makes an equally effective movie screen. we see) to heat, which cannot be seen (Planck Under some circumstances, neither is more n.d. 143). absorptive than the other. Planck's blurring of the distinction between The status of white as ideal reflector is as colorless and black probably rests on the dubious as that of black as ideal absorber. A reasoning that black is proverbially not a color. mirror is a more likely candidate for the title, Hence it can appropriately be associated with because what it reflects is more complete or heat, which, because it cannot be seen, is also,

92 Achromatic Colors and Mirrors

in its way, not a color. The Inter-Society Color reasonably assume that the blackness is the Council, sailing in an opposite direction but color of whatever lies immediately outside the navigating as erratically, decided to extend the universe. The sky might be regarded as, say, a of standard color names "by substitut- black curtain or sphere with a fire burning ing colorless for white," a usage that deprives outside. The stars are glimpses of that fire, seen us of a label for transparent conditions through holes in the curtain or sphere. (National Bureau of Standards n.d.a., 1). Twentieth-century creation myth paints a The word white is misused so often that more grandiose picture, raising new questions even Helmholtz or his translator floundered. without answering old ones. The universe, we We are told of Konig's discovery that some are told, has no "outside," a consequence of the individuals can detect no difference between curvature of space-time. The question, in that yellow, blue, and "a colorless mixture," case, is why this curvature, or light traveling whether colorless in the sense of black, white, through it, ought to result in a black color. We or transparent (Helmholtz [1909] 1962, have no reason to believe that interstellar space 2:404). is anything other than transparent. Yet the way Students in beginning painting classes, it looks, or the color it presents, cannot be reasoning similarly, have been known to ask explained by analogy with transparent objects what color paint to buy "to paint glass." The on earth, which transmit the colors of the question implies that, say, blue paint, yellow objects behind them. paint, and glass paint might exist. Although We are limited to just a few possibilities in this, of course, is not the case, the inquiry is not conjecturing about why the sky looks black. All wholly without sense. Colorlessness is a color return to the question of what causes black. condition: no object can be entirely blue and Prevailing wisdom identifies black as the entirely colorless or transparent at the same condition of default for human vision: the color time. Beyond this, many people have difficulty that appears when there is ostensibly nothing to understanding the phenomenological nature of see. Yet black does not appear to be transparency. categorically different from other colors, as one If an object (say, a pane of glass) can be said might expect if a switch were either on or off. to look transparent or colorless, it looks neither In another inconsistency, we see black objects black nor white. The colors seen when looking in the daytime and explain their blackness by at colorless objects are those of the objects assuming that they absorb all light. Interstellar behind them. This aspect of transparency is space looks black, yet ought to be a transmitter, difficult to reconcile with the dark color of the not an absorber, of light. Vacuums look sky at night. Vacuums, including the vacuum of transparent rather than black, at least in small interstellar space, ought to be the most ideally volumes. transparent of all objects. In theory, they If the black of the night sky is not to be contain little or nothing that could impede light understood as an accident of human neurol- or reflect it. ogy, the edge of the universe is what looks What and where, therefore, is the blackness black. In current theory the universe is finite. of the sky? If not to be understood as the color We are encouraged to regard it as a closed box, of what lies beyond the transparency, then the elastic and expanding as if it were a balloon colorlessness of interstellar space is dissimilar being inflated. The box is formed by the cur- to that of, say, sheets of glass. A primitive vature of space-time, a twentieth-century rein- regarding the night sky and reasoning by terpretation of the celestial dome imagined by analogy with transparent objects on earth might ancient peoples or of the spheres that enclose Achromatic Colors and Mirrors 93 the universe in Ptolemaic astronomy. stand that the reason for not putting it into use is The expansion of the box creates a that habituation to less excellent methods for presumption that anything outside it may be designating colors makes it difficult for human growing correspondingly smaller. But if beings to change their ways. Despite any aura of nothing can escape from the box, whether technological exactitude, however, no system anything lies outside is irrelevant. Is the box for naming colors by wavelengths would be itself transparent or opaque? If the curvature of workable. Too many serious impediments its surface is transparent (or cannot be seen), occur. The electromagnetic spectrum, for the blackness of the sky lies outside the box, a example, is devoid of wavelengths (therefore, of remarkable possibility that implies we can see wavelength notations) for colors in the redviolet what lies outside the universe. range and in other ranges as well. Because the box is closed, anything inside The two perceptual continua between red that bumps into its curvature is deflected and and violet (electromagnetic theory might at some later point enter a human eye. acknowledges only one) present another sort of This raises the question of whether the obstacle. If the wavelength notation for red is blackness we see could be something deflected 650 millimicrons, and that of violet is 400, 525 by this curvature, perhaps a previously millimicrons ought to be halfway between. But unrecognized component of light. If so, this we have no way of determining on which might salvage the theory that all colors are continuum, The median number might indicate accounted for by waves or particles that enter yellow-green, with a wavelength halfway an eye, most often after being deflected by between those for red and violet. Or it might as objects. This direction of conjecture implies validly point to red-violet, similarly between that the curvature of space-time looks black or red and violet but on an alternate continuum. behaves in a manner similar to any other black Correlation between hue variation and object. The curvature of space-time cannot, wavelength fails with both red-violet and the however, absorb light. That contingency blue-to-ultraviolet range. Even if this failure of requires us to assume that light could leak out correspondence could be overcome, most colors of an otherwise closed universe. cannot be explained in terms of hue alone. They therefore have no place on the electromagnetic Coding Colors by Wavelengths wavelength scale, no wavelengths and no The electromagnetic spectrum is schematic in wavelength notations. The excluded group its explanation of color because many colors includes gray, white, black, brown, pink, all are not on the scale. Nevertheless, the com- tints and tones, any color that deviates from monplace assertion that "color can be defined spectral purity (say, a slightly brownish yellow), in terms of light" passes as unimpeachable. fluorescent colors, iridescent colors, and From this follows the familiar suggestion, metallics, as well as such recent innovations as whether meant to be implemented or only the- dayglo colors and the colors of anodized oretical, that color names could be eliminated aluminum. The electromagnetic scale is a hue from language if light-wave notations were scale, and color has other dimensions in used in their place. Red, in a classification addition to hue. system of this type, would be described as a The electromagnetic wavelength scale does wave of 650 millimicron wavelength, blue, 450 not accommodate color value (lightness and to 500 millimicrons, and so forth. darkness), which creates impediments to notat- The system seems admirably exact if it is ing the achromatic colors by wavelengths. not examined in detail. We are left to under- Black, if attributable to absence of light, con- 94 Achromatic Colors and Mirrors

sists of an absence of wavelengths. One redness and greenness in two black swatches possibility is to notate it as 0, which skirts the seen from a viewing distance of 18 inches will problem that absence of light is an improperly be less noticeable from a distance of 150 feet. framed concept, irrespective of the merits of the We recognize black immediately and may under theory on which it relies. certain conditions also notice the direction in A more immediate problem is that the only which this black deviates from absolute black. I black that can be understood as an entire conclude that the names black and white have a absence of light (null notation) is an absolute visual function and refer to more than.just the black. If this absolute exists, it must be a single concepts of darkness and light. shade of color, not a range of colors. The large That reflectance can vary in degree is suffi- number of different blacks that we see, which cient to suggest that notating by wavelength are not absolute blacks, might be regarded off cannot be a complete system. Imagine that a blacks or near blacks, each properly labeled as bright blue and a black both reflect a certain black though not as absolute black. Each must wavelength of blue light but in different reflect, or be accounted for by, a few amounts. No way of differentiating between wavelengths rather than none. Thus zero (0) these colors is provided by a system that takes could not be a proper notation. note only of the wavelength of the reflected A system that notated the range of blacks or blue light. Unlike hue, color value (lightness/ near blacks in terms of the light they reflected darkness) and (saturation of hue) would be difficult to correlate, incidentally, cannot be correlated with wavelength. with the manner in which colors are presently named. Ivory black artists' pigment has been shown to be a weak reflector of predominantly White blue light (see figure 23-4). In a notational The notational problems presented by white are system based on wavelengths, ivory black similar to those for black but more complicated. would be classified with the blues, although we Absent from the solar spectrum and unassigned call it black. to any wavelength, white is regarded as a If reflectance of light is to be used as the mixture of all wavelengths, a less exact criterion in naming colors, any color other than quantifier than many. Because of this multiple absolute black can be classified as a very dark reflectance, a theory of how to notate a mixture variety of blue, green, red, or some other of wavelengths is required. The need similarly chromatic color. The Impressionist painters, exists in the case of any color that reflects more following this logic, contended that artists had than a single wavelength of light, say, a black no need for black paint, because black objects that reflected some red light and some blue. The could be represented without it. Although this is problem of developing a notational method for perfectly true, it fails to take full account of how mixtures is not easy. It has not been solved in, we process what we see. Two black color say, chemical nomenclature, which otherwise is swatches will be recognized as black before one marvelously intricate. is noticed to be, say, slightly greenish and the White, like black, is a range of numerous other slightly reddish. shades. All of these shades other than absolute Is the immediate recognition of the white might be regarded as off-whites and can swatches as black a matter of vision or of incline toward virtually any color other than acculturation? I suspect we must recognize it as white. Whites can be yellowish, bluish, visual, because it can be placed in a brownish, or grayish. The need, therefore, is phenomenological perspective. A respective to write notations for a mixture that will not

Achromatic Colors and Mirrors 95 always include exactly the same components. developed with the aid of a colorimeter, do not, The achromatic grays, because mixtures of however, measure light reflected from a black and white, again are a special case. They surface. They measure the relative amounts of would have to be notated as mixtures of noth- red, blue, and green required to obtain a match ing (no wavelengths) and everything (all for a spectral color for a statistically determined wavelengths), a similarly stupendous technical average viewer. problem. The spectrophotometer comes closer. A The problem of how to notate a mixture, or color sample is successively illuminated with how to take account of multiple reflectance, beams of given wavelengths, usually spaced at might appear to be solved in the familiar tables 10-millimicron differences in wavelength. The of tristimulus values (figure11-2).These tables, reflectance of the sample for each individual

Figure 11-2.Tristimulus values of the spectral colors (abridged). Adopted in 1931 by International Commission on illumination.

Wavelength X Y Z Wavelength X Y Z (millimicrons) (red) (green) (blue) (millimicrons) (red) (green) (blue)

380 0.0014 0.0000 0.0065 580 0.9163 0.8700 0.0017 385 0.0022 0.0001 0.0105 585 0.9786 0.8163 0.0014 390 0.0042 0.0001 0.0201 590 1.0263 0.7570 0.0011 395 0.0076 0.0002 0.0362 595 1.0567 0.6949 0.0010 400 0.0143 0.0004 0.0679 600 1.0622 0.6310 0.0008 405 0.0232 0.0006 0.1102 605 1.0456 0.5668 0.0006 410 0.0435 0.0012 0.2074 610 1.0026 0.5030 0.0003 415 0.0776 0.0022 0.3713 615 0.9384 0.4412 0.0002 420 0.1344 0.0040 0.6456 620 0.8544 0.3810 0.0002 425 0.2148 0.0073 1.0391 625 0.7514 0.3210 0.0001 430 0.2839 0.0116 1.3856 630 0.6424 0.2650 0.0000 435 0.3285 0.0168 1.6230 635 0.5419 0.2170 0.0000 440 0.3483 0.0230 1.7471 640 0.4479 0.1750 0.0000 445 0.3481 0.0298 1.7826 645 0.3608 0.1382 0.0000 450 0.3362 0.0380 1.7721 650 0.2835 0.1070 0.0000 455 0.3187 0.0480 1.7441 655 0.2187 0.0816 0.0000 460 0.2908 0.0600 1.6692 660 0.1649 0.0610 0.0000 465 0.2511 0.0739 1.5281 665 0.1212 0.0446 0.0000 470 0.1954 0.0910 1.2876 670 0.0874 0.0320 0.0000 475 0.1421 0.1126 1.0419 675 0.0636 0.0232 0.0000 480 0.0956 0.1390 0.8130 680 0.0468 0.0170 0.0000 485 0.0580 0.1693 0.6162 685 0.0329 0.0119 0.0000 490 0.0320 0.2080 0.4652 690 0.0227 0.0082 0.0000 495 0.0147 0.2586 0.3533 695 0.0158 0.0057 0.0000 500 0.0049 0.3230 0.2720 700 0.0114 0.0041 0.0000 505 0.0024 0.4073 0.2123 705 0.0081 0.0029 0.0000 510 0.0093 0.5030 0.1582 710 0.0058 0.0021 0.0000 515 0.0291 0.6082 0.1117 715 0.0041 0.0015 0.0000 520 0.0633 0.7100 0.0782 720 0.0029 0.0010 0.0000 525 0.1096 0.7932 0.0573 725 0.0020 0.0007 0.0000 530 0.1655 0.8620 0.0422 730 0.0014 0.0005 0.0000 535 0.2257 0.9149 0.0298 735 0.0010 0.0004 0.0000 540 0.2904 0.9540 0.0203 740 0.0007 0.0003 0.0000 545 0.3597 0.9803 0.0134 745 0.0005 0.0002 0.0000 550 0.4334 0.9950 0.0087 750 0.0003 0.0001 0.0000 555 0.5121 1.0002 0.0057 755 0.0002 0.0001 0.0000 560 0.5945 0.9950 0.0039 760 0.0002 0.0001 0.0000 565 0.6784 0.9786 0.0027 765 0.0001 0.0000 0.0000 570 0.7621 0.9520 0.0021 770 0.0001 0.0000 0.0000 575 0.8425 0.9154 0.0018 775 0.0000 0.0000 0.0000

96 Achromatic Colors and Mirrors

wavelength can be measured and compared Although I have been known by the same with the reflectance of a standard white for that name all the days of my life, a reasonable same wavelength. The spectrophotometer argument could be made that I have not always cannot directly measure reflectance for the been the same person. Over a seven-year various wavelengths of a mixed light, say, period, all cells in the human body are replaced, ordinary daylight. Its procedures are therefore and the cells in my body today are not exactly necessarily ponderous. the same cells present yesterday. Human beings, Imagine that a certain color swatch reflected like the colors we see, are not static. certain percentages of red, orange, yellow, Measurements are just the measurements of the green, blue, and violet wavelengths, and this moment, plotting points in a coordinate system was determined by a spectrophotometer. Why without halting the continuous changes would listing these statistics be more simple occurring in that system. than just calling the swatch white? The same I am not disturbed that a color swatch is not swatch will look slightly different in color on always the same. I am not always the same another occasion when the illumination has either, because this is the nature of changed. The swatch will require a new set of phenomenological reality. A hope of getting at spectrophotometric measurements. Indeed, the the facts once and for all, by measurement, is swatch needs an infinite number of sets of not realistic. No constant facts that might be measurements, to account for all possible ascertained about either colors or people exist. lighting conditions. Are we all to carry around The reason that we do not presently have spectrophotometers to determine the in operation a system for identifying colors by measurements applicable for the moment? wavelengths is that no such system can be Exactitude of measurement is not always complete, coherent, or even desirable. The necessary or practical, the exact reason that task is impossible in theory and would not be objects, including colors and people, have useful in practice. All colors cannot be names. I have been known by the same name explained in terms of wavelengths: no since birth. People who address me by my wavelengths have been assigned to pink, name have no need to know, say, my blood brown, silver, black, and other nonspectral pressure, number of brain cells or hemoglo- colors. All spectral hues cannot be explained bin cells, or other items that could be meas- in terms of wavelength, given, say, the dou- ured. Rarely would their having this ble location of blue and the missing wave- information on hand be useful. Even the length notation for red-violet, which in theory Internal Revenue Service, which wants to know ought to be a spectral color because it consists everything, wisely does not inquire about my of a mixture of red and violet, each of which cholesterol count. is a spectral color.

CHAPTER 12 Color Causes

All vision is colour vision, forLi it is only h by observing differences of colour that we distinguish the forms of objects. I include differences of brightness or shade among differences of colour. James Clerk Maxwell, Scientific Papers

I express variety of illumination through an understanding of the differences in the values of colors, alone and in relation.

Henri Matisse, Matisse Speaks to his Students

he notion that color is really light is has two names-the word light could then be too firmly entrenched in popular be- eliminated if color were used in its place. lief and the physical sciences to be dis- Because this bears on how causality (as in lodged by the less difficult assumption that "light causes color") is envisioned, imagine an light is really color. But if color cannot be observer looking at the sun and receiving a cer- completely explained as an inference from the tain visual sensation. The event is convention- experience of seeing light, it cannot be just a ally explained by assuming that an entity called name for perceived effects caused by light tight travels from the sun to the observer's eye. waves. The task of explaining light (a name for The explanation is circular. The only immedi- certain perceived color effects) as an inference ate evidence for the existence of the light, said from the experience of seeing color is less to have caused the experience, is that the fraught with difficulty. viewer had a visual experience the presence I shall play the devil's advocate by asking of light is intended to explain. On another what light is and by proposing that it does not planet, a simpler race of human beings might exist-or it exists only as a convenient fiction, a say that the sun, rather than light streaming label applied to color effects under some from it, causes any sensations experienced circumstances. In an ideal language-a code in when looking at the sun. They could argue that which every object has a name and no object they had discovered a greater truth, because

97

98 Color Causes Light

if light causes the sensation, the sun causes the Pushing one button rather than another may light. result in punishment or reward. Human beings I am interested in the diagrammatic manner probably imagined lines representing causal in which the event is typically imagined. The connections between objects long before those popular conception of light was entangled with lines came to be regarded, in some cases, as that of lines-including lines representing symbolic equivalents for rays of light. trajectories of particles-long before Maxwell Lines used for graphic purposes have their pictured the electromagnetic spectrum as a line own history, separate from the history of those with indeterminate ends. Drawing lines to marks that developed into systems of writing. represent rays, especially in halos, is a familiar At some unknown early date, human beings device in medieval and early Renaissance art as picked' out the constellations by imagining lines well as in drawings by children. Rays of light connecting one star to another in the night sky. (or their paths) continue to be depicted as lines They learned to imagine connections where in diagrams used as illustrations for modern none existed and to represent these connections texts on optics. as lines. Outline drawings in Paleolithic caves We think in diagrams, which breed their show a recognition that lines have a greater own confusion, reflecting what we believe potential than functioning as images of lines or rather than what we see. If asked to imagine of anything else that looks linear. Outline light traveling from the sun to an observer's eye, drawings, imagined in the sky or marked on most people visualize it in the form of lines cave walls, were the first abstract art, if only connecting the two. Or the lines might be because objects do not have outlines wrapping arrows, to indicate that the light has a direction. their surfaces like black wires. Human beings have long been accustomed to Outline drawings are usually regarded as imagining any relationship between objects in tracings of shapes, unconnected to the colors terms of a causal connection-of links that can be of objects. Calling them juncture drawings shown in graphic form. No observer of an would be thought peculiar. Yet we find (or observer watching the sky would have noticed invent) the outline of an object by looking at lines or arrows descending from the sun into the juncture between the aggregate of color anyone's eyes. spots associated with the object and those associated with its surroundings. In drawing The behavioral psychologist B. F. Skinner rays of light, outlines, paths, trajectories, or cir- believed "man's first experience with causes cuits, lines are manufactured that cannot be probably came from his own behavior: things correlated with other, similar black lines that moved because he moved them. If other things can be located in the natural world. Perhaps moved, it was because someone else was these imaginary lines were inspired by the moving them, and if the mover could not be horizon, the line separating earth and sky that seen, it was because he was invisible "(Skinner we see but cannot reach or touch. 1972, 5). The invisible "he" subsequently Whatever light is, it is not the lines so often became an invisible force. We draw diagrams to drawn to represent it. Nor is it the intimations of show what the invisible force is doing, and causal connection that the lines incidentally these diagrams become our private truths. convey. Above all, light is not the isolated ray, Causality of the push-pull variety presents a thin white line usually surmounted by a dove, too simpleminded a picture of the phenom- that we remember as a familiar accessory in enological world. Yet even laboratory animals Renaissance paintings of the Annunciation. recognize that actions have consequences. Rays of light seen in side view are not, we are Color Causes Light 99 told, what they seem. What is seen is the time" (Gribbin 1984, 190). scattering of the light by dust particles A curiosity of this modern picture of the suspended in the air. The ray itself is effectively relativity of time is that it raises the question of invisible from that oblique angle. whether light-a photon-is capable of movement in any ordinary sense. Because motion is an event that happens to forms in time, nothing Instantaneous Propagation moves in a formless void. Motion is similarly Until the seventeenth century, the propagation inconceivable without time, the dimension of light was thought to be instantaneous. The within which it occurs. Italian astronomer G. D. Cassini (1625-1712) The photon thought it moved from here to noticed variations in the elapsed time between there in no time at all. But it must also have eclipses of Jupiter's moons, according to concluded (if it is a logician-photon) that here is whether the planet was approaching or reced- synonymous with there. The logic of the ing from earth. Ole Roemer (1644-1710) syllogism is that the photon cannot change its recognized the cause of the discrepancy as the position. varying times required for light to traverse the What contemporary physics has brought us distance. Roemer made the first computation of to in its conception of light, space-time, and the speed of light in 1675, 140,000 miles per color-qua-light is the notion of a photon trapped second. The modern figure is 186,282 miles in a world of its own in which time has per second, a velocity that in theory cannot be collapsed, carrying space along with it if one is exceeded. continuous with the other. Motion is Although instantaneous propagation osten- paradoxically impossible and instantaneous at sibly has been discarded in contemporary the same time, This, at least, is how a photon physics, it may instead have been displaced. looks to a human being through the filter of A staple of relativity theory is that a clock (or relativity theory, without prejudice to how time) slows down according to the velocity human affairs might appear to a photon. with which it moves. At the speed of light, The speed of sound was surpassed during the time stands still. This is usually interpreted to 1950s at the cost of many lives. In retrospect, mean that time is a null dimension from, so to the barrier was just mechanical or aerodynamic. speak, the point of view of a photon. The speed of light is less approachable. A ray of light reaching the earth from the Equaling it implies hurtling into a world of edge of the universe may have taken, by our immobility, in which space is inaccessible (or manner of reckoning, fifteen billion years to instantaneously accessible) because of the arrive. If we had been able to ride on one of absence of time in which movement through it the light particles(a photon travels at the speed could occur. The four-dimensional time- of light because it is light), theory holds that space continuum is compressed to a Euclidean the journey would have been accomplished point, a one-dimensional (or nondimensional) instantaneously. As Gribbin pointed out, "for universe. The metaphysics is intriguing. a photon time has no meaning .... A photon The presumed cessation of time for a par- of the cosmic background radiation has, from ticle moving at the speed of light implies that our point of view, been traveling through the nominally four-dimensional time-space space for perhaps fifteen thousand million continuum of the photon lacks one of its years since the Big Bang in which the universe dimensions, The obvious parallel is color, as we know it began, but to the photon itself which, as we perceive it, is similarly devoid the Big Bang and our present are the same of a dimension: the colored images of the vis- 100 Color Causes Light

ual field have no depth. Can one dimension be cent objects. But this glow, because invariably a transformed into another? color phenomenon, does not require assuming Perhaps it can. Both relativity theory and the independent existence of an entity called quantum theory allow theoretically for travel light that is separable from color. through time. The technique allowed by Beyond that some objects glow, always relativity theory "involves distorting the fabric with a glow of some color, attempts to reason of space-time so that in a local region of about what light looks like lead to either para- spacetime the time axis points in a direction dox or a tangle of words. Illumination, for equivalent to one of the three space directions in example, is said to be the vehicle that enables the undistorted region of space-time. One of the seeing to occur. But illumination is also iden- other space directions takes on the role of time, tified as that which is seen, a merging of cause and by swapping space for time such a device and effect. Light acts upon photographic plates would make true time travel, there and back but cannot itself be photographed. Although again, possible" (Gribbin 1984, 193). light accounts for the formation of images, no Space-time can be distorted by a strong such thing exists as an image of light. Images gravitational field. are limited to those of illuminated or illuminat- I am less interested in time travel than in the ing objects. If we see by perceiving images, interchange of one dimension for another. this suggests we cannot see light or seeing it Interchangeability of dimensions other than is a radically different experience from seeing time is self-evident. In the case of a cube or anything else. other volumetric form, which measurement is I need not assume a table enters my eye regarded as its height, which width, and which when perceived. But when light is seen, the depth is arbitrary. Motion may be as impossible experience is explained by saying light has in the depthless world of the color spot as in the entered my eye. When a crowd convenes, the timeless world of the photon. The lines and aggregate amount of light that enters arrows used in diagrams mislead about the everyone's eyes causes no measurable drop in nature of motion on any type of planar surface. the general illumination. Yet if the same crowd They tempt us to assume that a color spot can were confined in a subway car in summer, their move laterally from here to there. How does aggregate body heat would affect the this differ, if it differs, from the spot temperature of the immediate environment. dematerializing in one place and materializing A river can be observed without touching it. in another? One answer to Zeno's paradox about But the only way to observe light or to col- the arrow, said not to move because it only lect information about it is to interfere with its rests consecutively at points in space, is that travels. A photographic plate, a measuring this is the nature of motion, especially on a pla- instrument, or an eye must intercept it by nar surface. blocking its path. The body of human knowl- edge about light is a collection of inferences about these impacts, an exquisitely restricted What Light Looks Like investigatory domain. Light is assumed to be To ask whether light exists as a perceptual continuously traveling through space when phenomenon amounts to asking for a descrip- nobody sees it or when allowed to pursue its tion of what it looks like. If not the spurious route without interruption. If we theorize that beam that is just illuminated dust particles, unseen light circles the universe to return to light may be, say, the glow around incandes- the spot from which it left, an impact of light Color Causes Light 101 with itself is implied. But the expansion of the most comprehensive of the three terms. People universe, by changing the location of the point failed to recognize that whenever they saw of origin, would cause the path described by what they called light, they were seeing color the light to be a spiral. However, light cannot effects. So they said light causes color, rather be seen when it is not being looked at, whether than color causes light. the nonlooking is by eyes or by instruments. Sometimes people said that vision causes Therefore, no way exists of demonstrating that color, which they explained to mean that color such interphenomena (in effect, nonphenom- is an accident of human neurology. Again, they enological phenomena) can occur. had cause and effect reversed. Had they said Unseen light, precisely because it cannot be that color causes vision, the mechanism would seen, cannot be elevated to more than an have been more clear. Less time would have assumption. Even the need for the assumption been wasted debating nonsensical questions can be eliminated by expanding Whittaker's that grew from these confusions about words. argument. If Newton's prism can be thought to Is this what occurred? The answer depends have manufactured the spectral colors, an eye on whether anything about light and vision is or instrument can be thought to manufacture verifiable yet not reducible to a color the light it sees or records. phenomenon, to the sensory experience of Never encountered and seldom discussed, seeing colors. In the case of light, a large unseen light seems destined to haunt superstructure of theory consistently tries to imagination, if only in the form of those separate it from color, to prove that one is not imagined lines streaming down from the sun: the same as the other. lines symbolizing that which cannot acquire To the ancient Greeks, light was not a name phenomenological reality until it encounters an for certain color effects they saw but was a eye or instrument. Unseen light is a conceptual separate entity attributable to invisible waves or necessity, hence its ghostly persistence. We particles. In modern theory, light has many need to believe it exists. It shores up the attributes different from those of color. The otherwise dubious proposition that light can be issue is whether those different attributes are shown to exist at all. If devoid of an unseen verifiable. Color is regarded as, say, unmoving, state, light exists only in the form of an impact, while light moves. What does movement mean less an object than a nonentity. in this case? Does theory tell us about what If light does not exist, a conceptual error occurs or just about what could occur? was made at any early date in human history. Relativity theory and quantum theory are People saw the phenomenon of color and mis- said to transcend many puzzles about the takenly gave it two names. Sometimes they nature of light by positing a world of events. called it light, and sometimes they called it We are not encouraged to inquire about the color. Compounding the error, the phenome- material nature, if any, of the objects, if any, non also received a third name: sometimes it that participate in the events. As in the was called vision, although everything visible Buddha's Fire Sermon, all that appears to be is just an array of color spots. substantial is discovered to be phantasmagoria. Sensing that a single phenomenon should We find ourselves in the picture as empty not have three names, people eventually tried spaces residing in empty spaces, a vision at to right the balance. But they went about it in least as plausible as that of unseen light forever the wrong way. They tried to eliminate the frustrated in its attempt to encircle a rapidly word color, though this is the most useful and expanding universe. 102 Color Causes Light

Because large pictures need to be balanced the chair is not interactive. But unseen light is a with simpler understandings, I am not fond of phenomenon of a different order. We are told A. S. Eddington's insistence that a table that light is affected by our looking at it, and the appears to be solid may, paradoxically, be experience is interactive. When I see light, my composed of emptiness: of subatomic particles eye stops the light from traveling further. The colliding in interparticular vastness. This may light acts on the eye by entering it; the eye acts be a serious misunderstanding of the nature of on the light by stopping it. paradox. In the absence of privileged coordinate These interactions suggest that light that is systems, all perspectives are coequal and never seen cannot be similar to light as we coexistent. Therefore, Eddington's table, if know it. The unseen light has never endured the made of bosons, fermions, or quarks, is at the metamorphosizing meeting with an eye or the same time made of wood, at least from our similarly interactive impact with a measuring human perspective. We assume too much in instrument. Having never, so to speak, crashed, believing we can set this human perspective the virginal rays must be as different from the aside at will. The theory that supports light we see as a speeding automobile is from a Eddington's argument is a theory that says we car crash. Furthermore, the usual aids to cannot. That a speeding automobile can be extrapolation are lacking. I can assess the reduced to probability waves is unlikely to be severity of a car crash because I know what cars more real or objective than that certain effects look like if not wrecked. Because nothing can may be anticipated if the automobile runs out of be known of the form light assumes before gas. meeting the receptor, the severity of the crash Modern theory, although Eddington would cannot be gauged. not have agreed, allows us to have our cake Once light is stopped, no way exists of and eat it too. We can believe in invisible starting it up again. It vanishes, apparently worlds of unseen particles and forces without without trace, into the barrier that stopped it. discarding the world that we see. Each is a Unlike water squeezed from a sponge, it cannot different perspective, a different coordinate be separated from the absorbing eye or system. And modern theory defends the valid- instrument. Although light is said to be radiant ity of multiple coordinate systems. If I were energy, once stopped, it no longer has (or is) the size of an electron, doubtless the world energy. The rays that entered Newton's prism would look different. But I am not the size of emerged, perhaps because the prism was an electron. transparent. Those that collide with eyes are If the universe is assumed to exist, no part of never seen again. They presumably catalyze the world of theory speaks to the question events in the pathway between eye and brain. of how we can separately verify that light What arguments can be given to prove that exists. Is light just another name for color? The light is an independent entity, not just another question reduces to whether light can be sepa- name for color? The historical argument says rated from its own colors, a separation that people have always talked about light. This presumably accomplished only when the light suggests the existence of some entity for which is not seen. they had invented a name. Yet incubi, the I am willing to risk believing that a chair is philosopher's stone, and the aether suggest we there when I am not looking at it. I have rea- cannot assume things exist just because people sons for making the assumption. I believe, until have talked about them. told otherwise, that the chair is not affected Judging by their more frequent occurrence by my looking at it. The experience of seeing in ancient literary works, words referring to Color Causes Light 103 light effects are of earlier vintage than color spot appears, there is said to have been no names. They more often cannot be traced to light. earlier words meaning anything different than The dependence of light on the presence of they do. I conclude that the first color effect color suggests analogies with motion, which people noticed, or thought seriously about, was depends on the presence of forms. Motion the difference between day and night. They cannot be seen in isolation from forms, and coined words referring to light and to darkness, visible light cannot be perceived in isolation an absence of light. Later, they made more from color. Each, in this sense, is dependent on refined observations about color differences in something else that precedes it. The the daytime, the differences, say, between red psychologist Rudolph Arnheim asks the and blue. This type of phenomenon was called essential question about illumination: "Is there color, with little or no recognition that day, such a thing, and under what conditions is it night, and all phenomena associated with light observed?" (Arnheim 1956, 297). and darkness were also color phenomena. I prefer the answer that light or illumina- Aristotle conjectured that colors were tion is never, strictly speaking, observed. It can created by a mixture of the opposites of be explained as an abstract idea around which darkness and light. He never carried this observations of color, or some kinds of obser- further by asking whether light/darkness and vations about color, are organized. Motion, in color are a single phenomenon. Perhaps by a similar manner, can be explained as an Aristotle's day the habit of distinguishing abstract idea around which certain observa- between light effects and color effects was too tions about forms or objects can be organized. firmly entrenched to be easily questioned. The statement that a chair has been moved, for Munsell, Ostwald, and all other modern color example, explains the observation that the theorists have had to address the issue more chair is now located a measurable distance directly. They realized color could not be from where it was at a previous moment. We explained through hue alone. Color cannot say the chair has been moved if, as a value-lightness and darkness-had to be taken matter of fact, chairs do not exist. Motion into account. They said nothing, however, presupposes the objects that move. about the relationship between, say, the light of Any statement about light that refers to its day and the darkness of night and appearance in the phenomenological world can lightness/darkness as color phenomena. By this be recast as a statement about color. The time it was assumed that physicists talked difference between night and day can be spec- about the nature of light, and color theorists ified by a catalog of the colors observed under considered only lightness and darkness of the respective conditions. At night, no color color. is seen other than that of the darkness. Dur- The inconsistency was noticed from time to ing the day, all colors are visible, including time, although the arguments were not carried black and gray. Without exception, describing to the conclusion toward which they pointed. differences in light means describing observ- The painter Henri Matisse and the physicist able changes in colors. James Clerk Maxwell pointed out that The concept of motion is as redundant as brightness and shade can be classified as vari- that of light, in that it just explains changes in ation in color, although traditionally identified the spatiotemporal location of forms. In vision, as variations in illumination.A light in the dark- these forms can be reduced to their images, to ness is seen as a spot of white color.If no white the array of color spots we see. Maxwell was 104 Color Causes Light

eminently correct in observing that in the visual sistency in modern theory: color is equated with field, form, motion, illumination, and all other light, which is said to move; yet we never say phenomena can be identified as color that color moves. phenomena. Because images are two-dimensional, color is more coherently understood as a Drawing Inferences About Light twodimensional matrix. Only the world we and Color touch has a third dimension, and touch is not For Bertrand Russell, "any definition of `red' vision. In three-dimensional terms, however, an which professes to be precise is pretentious intriguing question arises about light. Motion is and fraudulent" (Russell 1948, 260). We do, an event that happens to forms, and light is said however, make definitions. These definitions to move. We may have to recognize light as an incorporate popular but confused ideas about entity apart from color to explain such concepts the relationship between light and color, ideas as the movement of light. that are questionable inferences from the data The issue is whether anything is left if we thought to support them. The definitions and follow Max Planck in eliminating from the ideas eventually find their way into the concept of light anything, including its color, sciences where they assume an air of infalli- that is perceivable by the senses. In Planck's bility. reasoning, the invisible residue consists of the That red can be identified as light with a "real, objective phenomena." But where is this wavelength of 650 millimicrons is not really unverifiable residue? How can we confirm that true. The statement may be so loose as to be it exists if it lies entirely beyond the senses? meaningless. Red designates a broad class of Because modern theory presents us with a colors, because this is the nature of color photon entrapped in a world without time, and names. Irrespective of whether colors can be movement occurs only in time, we do well to correlated with wavelengths, red, because it is look closely at the concept of the movement of not one color, cannot be correlated with one light. If movement is transportation from here to wavelength. there, the movement of light is not only To accommodate its status as a range, red invisible but dissimilar to the movement of can be defined, in Russell's words, "(1) as any anything else. The question is whether we shade of color between two specified extremes reasonably call it movement in any conventional of the spectrum, or (2) as any shade of color sense. caused by waves having wave lengths between I suspect that the movement of light, these two extremes, or (3) (in physics) as waves originally called the propagation of light, is a having wave lengths between these extremes" figure of speech, words traditionally used to (Russell 1948, 259). Spectrophotometric describe a phenomenon that can be readings suggest that none of these alternates conceptualized in simpler terms. Can all are adequate either. A surface properly phenomena associated with the movement of categorized as red reflects varying wavelengths light be described without using the word in the red range (about 630-780 millimicrons). movement and without using the word light? It can additionally reflect quantities of orange, Perhaps change is a better word than movement yellow, green, blue, or light from any portion of in this case; perhaps color is a better word than the spectrum. light. If this or some equivalent reduction can The phenomenon of multiple reflectance is be accomplished, it may resolve one incon- consistent with commonplace observations Color Causes Light 105 when mixing paints. Adding a small amount of trum. If a crude approximation suffices, I will blue paint to a large amount of red paint provide generic names, such as red, blue, or presumably endows the mixture with some blue-violet. If a more exact specification is ability to reflect blue light. Yet the color of the required, I will look at the light and select a mixture, although not exactly the same red as color swatch that matches it as closely as previously, will likely still be properly classified possible. as red. Problems proliferate when the task is set Multiple reflectance is not the only forth in inverted order. If I show you a partic- argument against the assumption that red ular color, you cannot determine, without objects are ideal reflectors, either of red light or elaborate trial-and-error testing, the wave- of rays that match their own color. The color length or wavelengths that account for that pink, considered to be a type of red, can be color. At minimum, a spectrophotometer is mixed from red and white pigments. Because needed. These instruments are not entirely the white pigment nominally reflects all reliable and are less accurate than eyes when wavelengths, the pink presumably does the fine distinctions are involved. Although clas- same. This suggests a failure of the generality sical theory implies otherwise, most colored that red colors, as a class, can be correlated surfaces reflect more than a single wavelength with the range of wavelengths identified as red. of light. The visible light spectrum, which Particular red colors can be produced that ranges from approximately 400 to 700 reflect some light from beyond the red range or millimicrons, includes only variations in hue. from any portion of the spectrum. Mixtures of Variations other than hue occur among the ten light from the red sector of the spectrum are million different colors the National Bureau of insufficient if we want to replicate every Standards contends are individually variety of red. recognizable. It is not true that red-every individual The stronger generality for the circum- variation in the range we call red-can be stances is that light waves can be defined in defined in terms of light with a wavelength of terms of their color. No serious argument 650 millimicrons. What is true is the opposite occurs on the point that light with a wavelength of what is asserted. Light with a wavelength of of 650 millimicrons looks red. As if we had 650 millimicrons looks red in color. Given a fallen into the wonderland of Alice's mirror, particular wavelength, we can make what is conventionally asserted is the opposite reasonably accurate general statements about of what can be seen. Or, as often in the the color that light will look to an observer. explanation of color, eccentric assessments of If you specify any wavelength in the visi- visual phenomena survive because they ble light sector of the electromagnetic spec- integrate more easily with traditional beliefs and trum, I will tell you the color of light of that classical theories about the nature of radiant wavelength. I can easily look up the answer energy, the material world, and human in any diagram of the electromagnetic spec- experience. CHAPTER 13 Light as Symbol and Visual Metaphor I

Before the time of Newton, white light was supposed to be of all known things the purest. When white light appears colored, it was supposed to have become contaminated by coming into contact with gross bodies. We may still think white light the emblem of purity, though Newton has taught us its purity does not consist in simplicity. James Clerk Maxwell, Scientific Papers

he visual world can be thought of as given a name at an early date (its name is a continuum of color or an illusion borrowed from that of another entity), and ca used by light. Either mode is ade- names for light and light effects must be of quate for everyday purposes. Each excludes the earlier vintage. When color was named, it was other, implying a different context for thought of as the skin or superficial covering of understanding the perceptual universe. an object, to be differentiated from the essence Etymology provides clues about how the imagined to lie within. concepts evolved, why each assumed the forms The conception of color as a surface it took, and how human beings reasoned over phenomenon could reasonably have been based the centuries about the phenomena of the natural on the observation that many objects, especially world. life forms, have colors nn their surfaces that are The word light can be traced to the Teu- not continuous throughout their interiors. tonic root luh (to be light), similar in sound to Although interiors are no less colored than Latin lumen (a light) and lucere (to shine), surfaces, this easily escapes attention because Greek leukos (white), and Sanskrit rocate (he interiors ordinarily are not exposed to view. shines). Color, from Old Latin color, had an Celare can be traced back further to original meaning of a covering, from celare, to Greek Kalypteln (to cover, conceal) and San- cover or hide. This suggests that color was not skrit I~arana (concealing), roots for English

106 Light as Symbol and Visual Metaphor I 107 words as diverse as cellar, cell, and hell. night. Dark is from Middle English dark or The curious association of color and hell may derk, and Anglo-Saxon deorc, meaning dark. rest on an association between the sensory appeal The names of colors usually derive from the of color and hell as a final destination for those names of objects, and the generic term color who gave themselves to sensual pursuits. The similarly points to an entity other than itself. If suspicion that excessively bright color is heathen color was regarded as a covering (of objects), or barbaric is evidently of ancient vintage, this explains why it was associated with them persists in Goethe's writings, and is still with us and why so many color names are borrowed today. We like bright colors not because we from those of objects. Among names for hues, consider them genteel but because we have lost however, yellow and blue are unique. They interest in gentility. alone cannot be shown to be derived from the If, as it seems, interest in value distinctions is names of objects and are probably older than of earlier vintage than interest in color, this has other hue names. an understandable logic. People have practical Why would yellow and blue have been reasons for bearing in mind the distinction named or noticed earlier than other hues? The between darkness and light. Human beings, easiest answer is their value contrast, which unlike bats, distinguish objects more easily by resembles that between white and black or day than by night. The discovery of between lightness and darkness. Spectral fire-artificial light-is legendary and the remains yellow is light, like white. Spectral blue is dark, of artificially kindled fires are among the earliest like black. The pairs black/white and yellow/ surviving human artifacts. Torches, their blue occur together in the natural world and remains scattered in Paleolithic caves, originally might have attracted attention for this reason. served the purpose of the flashlights required The sky, black at night, is blue by day, today before braving the blackness of dark implying an endless cycle of permutations from cellars. black to blue. Sunlight, although usually called Practical considerations are not the whole white, is also described as golden or yellow. story. The visual difference between night and The bifurcated association is reflected in day is memorable because environmentally derivatives of aurum, the Latin word for gold. pervasive, more extreme than the differences In addition to aureole (the type of light effects seen in daylight between individual varieties of called a halo), these derivatives include oriole color. Seeing the greenness of grass is not the (the bird) and aureolin, the name of a pale equivalent of seeing nothing else. But at night yellow (or golden) color. the color of the darkness gives an impression of Although light is likely a more ancient being endless. concept than color, no record remains of the The etymological record suggests that light, date at which the earliest version of either word dark, night, and day attracted more attention came into use. Nor do we know why two words than color or individual colors and were given became current for what might have been names at an earlier date. Day, night, black, regarded as a single phenomenon in perceptual white, and dark, like light, cannot be traced to experience. But light and dark, which refer to earlier words meaning anything different than the difference between day and night, also they do. Day is from the Anglo-Saxon daeg, describe the quality in colors that became meaning day. Night is from the Anglo- known as color value. A model can be Saxon niht or neaht, similar to Icelandic nott, imagined in which human beings originally Danish nat, Gothic nahts, Greek nyx, nyktos, noticed lightness and darkness and later devel- or Sanskrit nakti and nakta, all of which mean oped names for the hues, while barely con- 108 Light as Symbol and Visual Metaphor I

sidering that both hue and value are ness are taught to children by ostensive characteristics of the same entity. The definition: by pointing at what the child ought phenomena originally called light and darkness to look at. could have been explained as aspects of color. The historical record follows the Color value is binary, although hue etymological record. Light traditionally has distinctions are not. The proverbial "difference been regarded as the phenomenological reality between night and day" is more extreme, by or primary: as percept rather than interpretation, experiential criteria, than the similarly cause rather than effect. Color, correspondingly, proverbial "difference between black and is secondary or superficial, an illusory skin that white." The difference is between may be unreal. homogeneous and nonhomogeneous color, not The body of symbolism associated with simply between one color and another. light, like names for light effects, is of ancient The colors that might summarize the vintage. Unlike , which is difference between night and day, if assumed to scattered, unsystematic, and probably more be the black of the darkness and the white of recent, the body of symbolic beliefs associated light, do not present themselves in orderly with light is too intensely focused to allow transition in the natural cycle. This may explain uncertainty about what it represents. The most why few names other than light gray and dark common association is with a supreme deity or gray have been coined for varieties of gray. As absolute. a color phenomenon in nature, the passage The sacred architecture of ancient and from the black of the night to the white light of medieval peoples provides a virtual day is interrupted by the roses, , and encyclopedia of the diverse religious ideas of its yellows of sunrise and . This may have builders, including ideas about the importance been the phenomenon that suggested to of light, of its absence, or of the interplay Aristotle or his predecessors that colors were between light and darkness. Among familiar mixtures of the opposites of darkness and light. art-historical examples, the Egyptian pharaoh Bright hues were seen in the sky at the time of Akhenaton's lost temple at Tell el-`Amarna transition from darkness to light. (Eighteenth Dynasty) is said to have had no The nonbinary nature of hue is reflected in roof. This allowed the worshiper to be bathed in figures of speech. "As different as black and the rays of the sun god. The device was white" is emphatic. "As different as pink and evidently a deliberate stylistic deviation from blue" is not and is rarely used. Among other the dimly lit interiors of earlier Egyptian etymological marks of the disassociation hypostyle temples. Abbot Suger's ideas about between color value and hue, lightness and the symbolic association of light with God are darkness are not names of colors. Even black regularly recounted as a preface to theories and white may be regarded as noncolors. Yet about Gothic architecture, especially those that none of these words can be explained in other refer to the large stained-glass windows that than visual terms. modulated the light entering cathedrals. An individual blind from birth can differen- Darkness, the opposite of light and there- tiate between night and day if taught, say, that fore its equal, has proved equally capable of night is the period when clocks indicate that suggesting religious experience. Examples the time is between 6:00 P.M. and 6:00 A.M. include the hypostyle temple, the Indian cave Vision is required for the more subtle task of temple, the womblike darkness of the confes- learning to distinguish between black and sional booth, even the practice of praying or white or between dark and light. Like color meditating with closed eyes, as if seeking an names, words referring to lightness and dark- inner light in the darkness. We rarely reverse Light as Symbol and Visual Metaphor I 109 the metaphor by wondering whether an inner definition, in the sense that darkness can be darkness should be sought in the light. exhibited. Not-color cannot be shown, and it is a Architecture has always been an art of more ambiguous, possible meaningless, spaces, therefore of light and darkness, the two concept. It implies the extinction not only of the visual conditions for space. The interior space visual field (the color domain), but of all of any building is a vessel into which light can memory and knowledge of it. enter or from which it can be excluded. But Although the usual name for not-light is buildings have exterior as well as interior darkness, if relative in degree it is called shade spaces. Symbolic ideas about light have or shadow. Not-color has never been thought to affected beliefs about where sacred buildings require a name of its own. The closest terms are should be located, not just how they ought to be colorless or transparent, which are not bona constructed. The recurring idea that the gods fide names for color conditions because they do speak from high places is said to explain why not describe surfaces of objects. They houses of worship were often elevated. Familiar denominate the interior (or essential) condition examples include the Mesopotamian shrine on of any object that allows the colors of the its ziggurat (intended to symbolize a mountain), objects behind it to be seen. The phenomenon is the Parthenon on the Acropolis hill, a an exception to the rule that I cannot see behind synagogue in a high place, or a Gothic objects in front of me. If the objects are cathedral towering above other buildings in its colorless or transparent, I see whatever lies town. behind them. I doubt we reach an entire understanding of Light and not-light appear as antitheses in this passion for heights without reflecting that the first chapter of Genesis. God's first act is what verifiably comes down from above is the creation of light, where previously there had light, stirring thoughts about who could have been only darkness, a darkness that may or may sent it. A shrine, temple, church, or synagogue not have been preceded by a more absolute in a high place is metaphorically the "first" "nothing." Adding to the uncanniness, light is among buildings in its vicinity to be presented as an entity that preexists radiance illuminated. We need not concern ourselves from luminous celestial bodies. Day and night, about whether the illumination is by the words darkness and light, appear on the first day of the most holy or by a more earthly light. (Genesis 1:3). The sun, moon, and stars are Goethe believed "darkness and light have absent until the fourth (Genesis 1:16). A eternally opposed each other, one alien to the perennial question, discovered by clever other" (Matthaei 1971, 126).The opposition is children, is what the source of the light could more likely rooted in language than eternal, a have been during the first three days of literary concept grown beyond its modest creation. visual foundation. Terms such as enlighten- One answer is that Scripture, whether it ment, elucidation, illumination, and lucidity originates or just perpetuates the paradigm, equate light with understanding and grace. pays scant attention to the nature of perceptual Darkness is the implied opposite in each case, experience. Biblical personages are identified though it has become archaic to refer to the by reciting their genealogies rather than by unenlightened as the benighted.A major differ- descriptions of what they looked like. The ence between light and color is the greater ease orientation of the Bible is nonvisual. Its authors with which light can be envisioned in terms of or compilers were indifferent to the sensuous parity or polarity. If a foundation for the delight of telling about what can be seen. Both impulse ought to be sought in the phenomeno- Old and New Testaments are devoid of logical world, not-light is capable of ostensive Homeric rosy-fingered dawns, although similes

110 Light as Symbol and Visual Metaphor I

and other literary devices abound. A bride may itself that lies behind it. be compared to a garden. A report about what a Language rarely transcends religious bride (or a garden) looked like is less likely to metaphor, a form of double-speak inherited be encountered. The limited worldly splen- from its past. Musings about the meaning of dor evoked is that of perfumed odors and - light remain a perennially popular, though ous music. Little is said about shapes and overworked, form of poesy. Among numerous colors. borrowings in The Waste Land, T. S. Eliot The biblical injunction against graven reworked a passage in The Aeneid in which images is consistent with this context. Those Virgil describes the light of torches that hang who passed along the prohibition by setting from the ceiling in Dido's palace (Eliot, Waste down the words of the Old Testament show Land, note to line 92). But in "Virgil and the little interest in visual experience. Nothing Christian World," Eliot complained that Virgil, excites them in the task of depicting or unlike Dante, sees light only as a physical entity capturing, in words or pictures, how anything and overlooks its symbolic possibilities. An looks. Their priorities, or their morality, lay absence is noted, in The Aeneid, of "lume, and elsewhere. all the words expressive of the spiritual Within this ambience, the opening of Gen- significance of light" (Eliot 1957, 147). esis, with its striking imagery of light and dark- Eliot apparently believed that concern for ness, may seem a false note. Light and dark, this significance was uniquely Christian, or because the blind see no difference between Judeo-Christian. Yet Christian symbols are them, may seem unambiguously visual. Yet the often of pre-Christian vintage, as in this case. In light that opens the first chapter of Genesis is the second book of The Aeneid, flames appear an unusual light. We do not discover this light around the head of the infant Ascanius. The subsequently illuminating a world to be adored flames are taken as an omen when they are for its magnificence or for the wonders of what discovered to be illusory. The passage is often we see. The universe of the Old Testament is cited as an early example of the image of the painted in unrelieved black and white. The halo, effectively of the symbolic use of light. differences between the path to be followed We do not know who originated the idea that and that to be avoided are as great as those light could be a sign or an omen, that it could between, so to speak, night and day. The faith- have purposes beyond a utilitarian purpose. The 4ful are enjoined to love God, who cannot be Star of Bethlehem was read as a supernatural seen, and not the world, which is visible. This omen. So was Halley's comet, when it appeared orientation does not encourage valuing of the in the sky in A.D. 1066, just before the Battle of visual sense. Hastings. Disjunction from perceptual reality (from what can be seen) is the point of the supranor- mal light of the first three days of Genesis: the Light as God, Dark as Void Ur-light that has no identifiable source other Equating God with light cannot be called than God. The imagery intimates that light can anthropomorphism in the most rigorous sense. be envisioned as other than the physical It differs from conjuring up a mental picture radiance of incandescent objects. Light can be of an old man with a white beard. But it satis- read as a symbol. As such, it acquires a value fies a similar need to recast the imperceivable beyond its inherent worth as an object. At a in comfortingly familiar form. A parallel com- much later date, Saint Augustine recycled the pression allows darkness, lightlessness, or motif in admonishing that the sun is only a blackness to function as a figure for the void: reminder of its maker, of the light greater than the emptiness from which we hope God will Light as Symbol and Visual Metaphor I 111

save us. This abyss is to be distinguished from darkness enshrouds what escapes us. The the emptiness that is God. The metaphor metaphor, handed down from the past, long ago suggests, intriguingly, that one can be degenerated to the tediously platitudinous. We differentiated from the other. have every reason to understand darkness, at Visualizing an abyss filled with darkness is least as well as we understand light. Even not difficult, although it may be too easy. The though most of us hope to be able to sleep once immensely popular "In victus," by through most of them, the total number of William Ernest Henley (1849-1903), is one of terrestrial nights is exactly equal to the total many literary works that project a similar number of terrestrial days. epiphany. The radiance of spirit is to dispel an Night is different from day. But why is it otherwise all-encompassing not-light. regarded as mysterious? I suspect what unsettles us is the color field experience, seeing a single homogeneous area of color Out of the night that covers me, uninterrupted by other colors. Black as the pit from pole to pole, I Light, shadow, and the interplay between thank whatever gods may be them acquire ornate symbolic overtones for For my unconquerable soul. adults. The infant may lack a natural sense of ……………………. these subtleties or even an interest in the I am the master of my fate; phenomenological foundation from which they I am the captain of my soul. arise. Arnheim pointed out that young children typically do not include light and shadow Darkness disappears with the arrival of light, relationships in their art (Arnheim 1956, 310). as human experience suggests. For Henley, as When taught to do so, they acquire an interest in for others, metaphysical solace lay in imagining modeling with light and shadow that replaces the void as a dark night inflated to the earlier emphasis on color. macrocosmic dimension. Because children also need to be taught the If the metaphysical abyss is to be a difference between right and wrong, I suspect meaningful concept, we need to imagine it as that what is being taught, in either case, is a more than just a big pot of black paint. Based habit of thinking in terms of opposites, whether on a visually insensitive, although traditional, light/shadow, black/white, or right/ wrong. The reasoning about the color black, Henley's child requires an understanding of oppositeness metaphysics is atrocious in its own right. The to comprehend the process of reasoning, based difficult task, before which imagination falters, nominally on the discipline of logic. We all is to imagine a void filled with nothing, a figure learn, and often need to unlearn, the strict for universes beyond human understanding. In distinction between true and false that allows no the face of the abyss, talk of mastery or room for what is more or less true. The price of illumination is empty, and even darkness and this education is sensitivity to the continuum of light may be meaningless concepts. If any of the perceived world, including the world of us, as Henley optimistically imagines, were visual experience. truly the masters of our fate, I doubt we would I do not mean that the concept of light and voluntarily choose to die. dark is so narrow as to be no more than hand- Whether or not to trivializing ends, the maiden to logic and morality, to the imagined counterpoint of light and darkness is familiar polarity that inspires us to assert a distinction poesy for marking the limits of human com- between order and disorder. But it rarely rises prehension. That which men and women far above this primary impress, even in the understand is figuratively illuminated;a deeper visual arts. In painting, modeling with light and 112 Light as Symbol and Visual Metaphor 1

shadow is inseparable from the representation invisible world might exist. of three-dimensional form on a If a high degree of symbolic significance two-dimensional surface. The forms of the was attributed to light and darkness at a corporeal world are what light and shadow remarkably early date, the relatively late reveal or conceal. appearance of representations of light in paint- Rudimentary modeling with light and shade ing may seem out of character. We cannot can be seen in some Paleolithic cave paintings, assume, however, that paintings always show notably at Altamira. But painters of the Roman us what people think is most important. The Empire, possibly preceded by the Greeks, were concept of light, like the concept of soul, was the first to exploit the device in art consistently. excessively overburdened with metaphysical Pre-Roman (or pre-Greek) painters rarely complexity. It became intimidating or over- represented light and shadow. Indeed, it plays a whelming. A belief that a phenomenon is sig- subordinate role in virtually all art except nificant need not lead to immediate answers European, and European-influenced, art from to questions that are critical in the art of paint- the early Renaissance until the late nineteenth ing. One question is how the phenomenon can century. This cannot be because shade or be represented on a two-dimensional surface. shadow was unimportant to, or unnoticed by, Other questions are whether it can be other peoples. More likely, an excessive represented at all and how its depiction can importance was attributed to it. Long before be integrated with other artistic goals and Robert Louis Stevenson penned "I Have a Little limits. Shadow," folk tales from all over the world Even today, some light phenomena are equated the human shadow with the human rarely represented, because essentially soul. A person's soul could be stolen by seizing unpaintable. Although the opening chapter of his or her shadow. Genesis is famous, Michelangelo's illustration In Western painting, concern for the natu- of it on the Sistine Chapel ceiling is ralistic representation of light and shadow exceptional. God's separation of the light from waxes and wanes with interest in perspective the darkness is awesomely nonvisualizable, drawing and the related device of foreshorten- therefore rarely depicted in painting. I cannot ing. These tools configured the picture plane in imagine what the event would have looked like, terms of the polarities of near/far and here! and no analogies come to mind that might help. there. Without them, the illusionist tradition in The deity of the Sistine Chapel ceiling, poised Western painting could not have developed or between a black cloud and a white cloud (the would have assumed a different form. darkness and the light) waves his arms as if he For Alfred North Whitehead, discovery of were a magician performing tricks on a stage. how to make "true-to-life" paintings Michelangelo's brave effort falls far short of the encouraged the development of modern metaphysical grandeur of the biblical words. science, which profits from similarly close Later illustrators of the Bible by and large left observation of natural phenomena. Whitehead's that passage alone. point has merit, though it is too idealistic. A Starlight is another painterly enigma. Dante long time has elapsed since the scientific study is famous for evoking it by ending each cantica of light implied a disinterested contemplation of Commedia with the word stelle. But of sunsets, rainbows, or other natural light Rembrandt, that great student of all manner of phenomena. Humanity has sought its scientific light effects, left no depictions of the stars at understanding of light, as it sought its gods, in night. The scintillation (twinkling) of the stars the invisible world. The role of the natural is temporal and therefore cannot be painted. world was only to disgorge evidence that this Dots of white paint displayed on a black cur- Light as Symbol and Visual Metaphor I 113 tain fail to evoke the awe that the sky at night spective, codified by the Florentine architect has traditionally inspired. Van Gogh painted the Filippo Brunelleschi (1377-1446) and night sky with its stars several times. I am developed further by Durer and others, were tempted to wonder if this is because he was once regarded as an exciting innovation, a high mad. Starry Night, that justly famous endeavor, technology of their day. The laws of cuts the Gordian knot by abandoning perspective revealed that although no object has naturalism; the lavish swirls provide, instead, a constant perceivable size or shape, the one artist's objective correlative for starlight. relativity of vision has a logic. Extension in the visual field (size and shape) Light and Perspective is relative, in particular ways, to a particular Diagrams in textbooks on optics show how viewer at a particular time and place. Those rays of light pass through lenses. These ways can be determined by an adaptation of diagrams look remarkably similar to geometry or by borrowing the methods of perspective drawings, though the vanishing construction used in geometry. In point is replaced by a focal point. Either type of foreshortening, the application of what is picture is a geometrical construction. What is essentially a calculus enables an approximation diagrammed is the manner in which rays of of the perspective of those forms (for example, light move from a point of origin to a terminus. the human hand) too complicated to be entirely If the rays did not travel (radiate) in the manner reduced to combinations of simple geometrical they do, there would be no such thing as shapes. perspective, the progressive diminishing of the Virtuoso displays of skill in perspective visual size of any object as it recedes from a rendering, and in the foreshortening of the forms viewer. of the human figure, are common coinage in In art, perspective drawing and modeling Renaissance art. Examples include Botticelli's with light and shade supplement one another. Adoration of the Magi (Washington, National The aim of perspective drawing (or mechanical Gallery); Raphael's School of Athens (Vatican drawing) is to present an accurate Palace, Stanza dells Segnatura) and Marriage of twodimensional projection of the perceived the Virgin (Milan, Pinacoteca di Brera); shapes of objects in the physical world. Mantegna's The Dead Christ (Milan, Pinacoteca Although elsewhere we doubt the reliability of di Brera) and St. Jerome Led to Martyrdom the visual sense, only this accurate projection (Padua, Oretari Chapel); Perugino's Giving of of what we see is regarded as a correct or true the Keys to St. Peter (Vatican, Sistine Chapel); picture. To make a drawing of, say, a building Jan van Eyck's Madonna with Chancellor Rolin in proper perspective is held to be synonymous (Louvre); and van der Weyden's St. Luke with drawing the building correctly, an Painting the Portrait of the Virgin (Boston, assumption rarely disputed even by those who Museum of Fine Arts). question whether "correct drawing" is In these and other works, perspective necessarily the aim of art. drawing assumes a fixed viewpoint that is an Students in art schools are known to be artificial convention. The question remains generally unenthusiastic about the ubiquitous open whether perspective really bears a one- required course in mechanical drawing, per- to-one correspondence with visual perception. haps a reaction to the moralistic tenor. The Human beings do not typically see from either implied distinction between a right and wrong a single eye or a fixed viewpoint. The eyes, of way to draw is coupled with a demand, in most which we have two, continually scan. And the cases, for a nearly unattainable level of world is neither petrified nor a mirror image of neatness and precision. Yet the rules of per- ideally frozen forms that transcend its flux. CHAPTER 14 Light as Symbol and Visual Metaphor

Such also before others are accepted into heaven, and are among those there at the center, because they are in light more than others. Emanuel Swedenborg, Heaven and Its Wonders and Hell When Swedenborg's message was revealed to me . . . it was as if light came where there had been no light before; the intangible world became a shining certainty. Helen Keller, Introduction to Heaven and Its Wonders and Hell

erspective drawing and the rendering of A more modern conception of perspective light and shadow can aim for illu- occurs in Parmigianino's youthful self-portrait si onism, which we often equate with in a convex mirror. What we see depends on the visual truth. Either can also become an end in focal length of the lens through which the itself if pushed to extremes for dramatic effect. perceived light passes or of the mirror in which Playing with foreshortening and perspective an image is reflected. The artist's large hand and began in the early Renaissance, and some artists much smaller face prefigure the images seen in aspired to more than just the obligatory fish-eye lenses, Weegee photographs, Picasso landscape seen through an open door or paintings. window. Michelangelo used extreme Although regarded as distortions of the way foreshortening in the figures of nude boys on the world looks, images in curved mirrors and the Sistine Chapel ceiling. Pozzo used extreme through fish-eye lenses are the way the world perspective in The Apotheosis of St. Ignatius. looks in those mirrors and through those Both works are tour de force art-historical lenses. No object has a normal shape or color markers. They show that painters, with that remains the same under all conditions. perseverance, learned to represent any object Colors vary with illumination, and shapes vary from any viewing position in "true perspective." with spatiotemporal location. What we see is

114 Light as Symbol and Visual Metaphor 11 115 affected by the color balance of the light tive and in any foreshortening, the skill had illuminating the scene and by the path the light become a cliche. rays travel to reach the eye. A penguin ten feet The motif of enveloping shadows, which from me does not have an unchanging size and conceal forms light would reveal, grows with a shape even for me. Size and shape vary life of its own. We can trace the iconographical according to the lens, or combination of lenses, history of forms hidden in semidarkness through which I look. The limit is that no lens through the paintings of, say, Masaccio, can make a penguin look like a kangaroo. Leonardo, Caravaggio, and Rembrandt. In El When used to create styles of painting that Greco's paintings, shadows that are unusually are not naturalistic or illusionistic, perspective dark (nearly black) play against illuminated drawing and modeling (light and shade) areas that are unusually light (nearly white). diverge, evoking different responses. Extreme The colors of objects are revealed through the perspective effects amaze viewers or are middle tones. El Greco forces both light and regarded as distortions. They rarely seem shadow. Forms bend and twist from the stress. metaphysical or mysterious. We attribute Although El Greco disliked Michelangelo's metaphysical overtones to paintings that paintings, his own have a similar tension include extremes of light and shadow. Artists achieved in a different manner. arrived at an interest in these extremes by a Why do we read transcendental intimations series of steps, by solving one after the other into strong light and shadow effects? Artists what are now called painterly problems. They who rely heavily on light and shadow are not saw certain challenges, met them, and moved always admired more than others for that on to others. reason (figure 14-1). But painters credited with Late Gothic and early Renaissance painters revealing profundities about the human soul are struggled to master the representation of the usually those who create shadows that are read three-dimensional world on a two-dimensional as mysterious, Leonardo in St. John the Baptist surface. They wanted to understand forms, (Louvre) and Rembrandt are prime examples. which for them meant the anatomical forms of So is the Abstract Expressionist painter Mark the human body and the perspectival forms of Rothko. Fascination with chiaroscuro, distant landscapes. The edges of forms are tenebrism, and all manner of light and shadow precise in, say, van Eyck's paintings, either effects is enduring. because he liked the effect of exactness or As visual metaphor, the play of light against because he wanted to show how carefully he darkness can be read as confirmation that the looked at all the details of what he saw. world is less a continuum than an interplay of Almost as soon as representing forms opposites, reducible to what is easily clearly and exactly was discovered as a signi- understood. In the cosmic melodrama, a replay ficant problem, forms began to get lost again. of Zoroastrianism, Ahura Mazda, the god of They disappear into deep shadows in many light, enters the lists to joust with Ahriman, lord Mannerist and Baroque paintings. Artists had of the darkness. The outcome of the battle may not given up on the goal of understanding unveil the mysteries that never were spoken. three-dimensional forms. They had passed Having institutionalized the idea of beyond the goal and no longer saw it as a rele- oppositeness (which unfortunately is insepara- vant painterly problem. At a date when any ble from competitiveness), we temper it only trained artist had learned to draw anything, insignificantly by allowing that familiar further including the human figure, from any perspec- possibility: perhaps all opposites can be recon- 116 Light as Symbol and Visual Metaphor II

Figure 14-1. Light and shadow: opinions over the centuries.

Stendhal: "The great fault of the French School of by the play of light and shadow, and the merely painting is the total lack of chiaroscuro." picturesque projections of the current fashion, will be John Ruskin: "Now if there be one principle, or inclined to flout me when I say these are a jargon and secret more than another, on which Turner depends no tongue." for attaining brilliancy of light, it is his clear and Leon Battista Alberti: "I should hardly ever think of exquisite drawing of the shadows. " a painter as middling-good who did not know exactly Eugene Delacroix: "Painters who are not the effect light and shade produce on every surface." produce illumination, not painting." Roger de Piles: "The knowledge of lights and shades, Paul Cezanne: "Light does not exist for the which painting requires, is one of the most important painter." and essential branches of the art. " Heinrich Wolfflin: " [Leonardo's Judas] is the only Charles Alphonse du Fresnoy: "The shining one who sits quietly with his back to the light, and eminence of Corregio consists in his laying on ample whose features are therefore in shadows." broad Lights encompassed with friendly Shadows, Horatio Greenough: "He whose eye is tickled and in a grand style of Painting, with a delicacy in the arrangement of the Colours."

ciled eventually. They may paradoxically be tury, like the nineteenth, brought famous one, although not one. advances in the understanding of light in the Through the writings of Ogden Rood, who physical sciences. It was the age of Newton and popularized the ideas of Maxwell and Hooke. It was also the age of Caravaggio Helmholtz, some of the French Impressionists (1573-1610), Artemisia Gentileschi (1597- and neo-Impressionists came to regard 1651?), Bernini (1598-1680), Rembrandt themselves as researchers into the science of (1606-69), and Vermeer (1632-75). Interest in color and light. Seurat, and for a time Pissarro, light characterizes both painting and the saw themselves as researchers in the most physical sciences at that time. It was part of a strictly scientific sense. Whatever respectability broad social paradigm rather than proprietary to they thought science could lend to art, A. S. any single discipline. Eddington later returned the compliment. He Rembrandt was sixty years old when New- explained his conception of the subatomic ton performed his prism experiments. Rem- world by comparing electrons and protons to brandt's passion for finding evidence of God's the color spots in an Impressionist painting: perfection in the natural world was probably as small discrete units that made a whole but were ardent as Newton's (and, later, Maxwell's) not connected with one another. is said to have been. Rembrandt may have No reason exists to suspect that Monet or been less assured in advance of what was Seurat had premonitions of quantum theory. supposed to be discovered. What was found that Human beings recycle one another's intuitions, was new was an absence of hierarchical rank- including intimations so widely shared as to ing among human souls, a finding that predates be societal or collective. The seventeenth cen- by a century the French Declaration of the Light as Symbol and Visual Metaphor II 117

Rights of Man. Old baggage, passed on by To defend the idea that Vermeer is as great Caravaggio, was the habit of viewing light and and profound a painter as Rembrandt is not dif- darkness as extremes. ficult. But the literature suggests Rembrandt is Most of us have no basis for understanding more often praised for offering psychological the proclivities of those who had no electricity insights into the human condition. Crucial to light their homes. We rarely sit indoors in the expectations are satisfied for viewers: that the dark. Newton sat in a dark room watching a opposites of darkness and light mark the single ray of light pass through a slit and a bounds of experience, that human beings are prism before it fell on the wall. Goethe later the measure of all things, the noblest of crea- complained that this was not a reasonable way tions. We remember light and shadow from to study colors. Rembrandt painted illuminated Rembrandt's paintings, along with human figures emerging from darkness, as if separated faces. From Vermeer's, we recall fragments of by a theatrical spotlight from the surrounding another sort: the reflection in a brass dish of shadows. Whether Rembrandt painted under the colors of an Oriental rug, the whiteness of low-light conditions or just imagined them, the light coming through a window with gold device is an inspired compositional tool. mullions. Human beings (including viewers of paint- If a catalog were complied of the visual ings) look at light rather than darkness if given configurations that continue to engage viewers a choice. This is not because light symbolizes most, the image of light playing against shadow God, but because light provides better condi- might head the list. Like conventional tions than darkness for seeing colors and struc- perspective, this type of imagery plummeted in tural details. The tendency to look at light can importance after the Impressionists and be called instinctual, though I prefer to call it Cezanne. From the 1950s onward, it reappeared utilitarian. In Rembrandt's paintings, a catalog in new forms. A list of Abstract Expressionist of what is illuminated (or represented as illu- painters popularly credited with proferring minated) is a catalog of elements the artist metaphysical profundity would lean heavily wanted looked at, or thought were composi- toward those who use dark, murky, or shadowy tionally important. colors, especially if unrelieved by lighter tones The paintings of Rembrandt's maturity that might suggest brightness or light. Mark mark the end of the road for an idea. Vermeer's Rothko and Ad Reinhardt are examples. interest in light was different from Rem- Great art has been created in the past. But brandt'sand prefigured the French Impres- this is not the past. We need new metaphors to sionists. His paintings are relatively free of replace old ones that have been recycled too shadows.Light is a fluid flowing through every often. The unknown is not a big bucket of black crevice of the environment and uniting all paint. Nothing profound or significant is said parts. By bringing out the colors of objects, the about the human condition by comparing it to light invites consideration of those colors darkness and light or black and white. in their own right. The burden of inscrutable New metaphors will be found by looking at darkness has fallen away, along with its sym- the discredited parts of human experience, bolic associations. The eye, which finds a light parts brushed aside because they have labels in the darkness in Rembrandt's paintings, like delusion, illusion, random, accidental, travels amore complicated route through the unreal, or subjective. Color and other items in many colors of Vermeer's illuminated envi- this class (emptiness, nothingness, random- ronments. ness, and so forth) do not comfortably fit into 118 Light as Symbol and Visual Metaphor 11

our present system. But we have no other seeds and in terms of the picture plane. This for a new system, a new way of understanding congruency is lost if the head is rotated to, say, the world and ourselves. a profile view. If the halo is still "behind" the head, it is no longer clear what behind means. One of two solutions for the profile view Halos is illustrated by the angel in Botticelli's Annun- Halos provide a striking example of the ciation (Uffizi). Its halo is a golden pie plate symbolic use of light in the visual arts. The seen from the side. It lies behind the head in, several ways of representing them constitute an so to speak, real space-time. Because the encyclopedia of ideas about what light looks angel's head does not overlap the halo, the like or how it ought to be conceptualized. Halos halo is not "behind" in terms of the picture appear in religious art as a radiance around the plane. heads of the saintly, a convention for indicating In the other possibility, the profile face the touch of the godhead. The larger mandorla overlaps a frontal halo. The effect is as if the or body halo appears behind Christ in medieval disc had transported itself through real space art, but is not unique to the art of Western time to relocate "behind" the head in terms of Europe. Mandorlas also appear in the picture plane, though not spatiotemporally. representations of the Buddha. Neither option can be shown to be more correct Each variation resists explanation in terms of than the other. mimesis, the imitation of nature. Halos, like As might be expected from the ambiguities dragons and angels, are not seen in the natural that arise in profile view, disc halos rarely or world, though they have their own iconographic almost never appear on heads seen from the history. Halos appear in painting long before rear. But Giotto, in his Last Supper (Padua, artists had acquired illusionistic skill in the Arena Chapel), valiantly takes on the problem. rendering of illumination. Many do not look The attempt was, so far as I know, not repeated. like realistic representations of light. But they If depicted behind the head in real space-time, were correctly read over the centuries as visual the disc halo overlaps the head on the picture metaphors for it. plane and obscures it. If behind in terms of the In medieval and early Renaissance art, the picture plane, head overlaps halo. This halo is often a gold or yellow disc, situated unfortunate image is indistinguishable from behind a head seen frontally. Although light is that of a human figure confronting the disc of called white, disc halos are rarely this color. its own halo. The iconographic habit of placing The frequent use of gold leaf suggests, as in the all participants at the Last Supper (or all except golden skies of Byzantine icons, an intent to let Judas) on the same side of the table, all facing opulence carry the message that could not be the viewer, sidesteps the problem of depicting conveyed by illusionism. Gold, the color of haloed heads from the rear. neither light nor the sky, transforms either into The lines radiating from Mary's head in intimations of a world different from this one. Rogier van der Weyden's Annunciation (New Derivation of the word aureole (halo) from the York, Metropolitan Museum of Art) illustrate Latin aurum (gold) suggests halos may have an improved type of halo. Although each line been thought of as golden before they were is meant as an individual ray of light, the painted that color in art. imagery is symbolic rather than illusionistic. When the human head is represented fron- Rays of light, though linear, are not lines. Bot- tally, with a disc halo behind it, the halo is ticelli, who may have been apprenticed to a behind the head both in spatiotemporal terms goldsmith in his youth, develops further vari- Light as Symbol and Visual Metaphor 11 119 ations. The haloes look as if they were made of able to sense it in another. gold wire or filigree work in his Virgin and Leonardo's degree of complexity is rarely Child with Angels (London, National Gallery), approached. But later artists developed further and Madonna of the Book (Milan, variations on halos. In Titian's Savior of Hor- PoldiPazzoli). telano (Prado), rays of light shine from Christ's The halo floating above the head of the head to form a cross. In El Greco's Saint Martin angel of Leonardo's Annunciation (Florence, and the Beggar (Washington, National Gallery), Uffizi) is a tour de force of perspective curving clouds form a ring-shaped halo behind rendering and graphic analysis. Seen in side the saint's head. In Adam Elsheimer's Baptism view, it consists of spokes radiating from a of Christ (London, National Gallery), Christ's center, a combination of disc halo and ray halo. halo is created by four airborne cherubim In the even more splendid playings of The Last joining arms to form a circle. Supper (Milan, Santa Maria della Grazie), The question of who ought to have a halo Christ is depicted without a halo. But Leonardo has been answered differently in different has implied its presence in four ways. More societies. Byzantine emperors and empresses, exactly, he reduces the halo to four qualia, each unlike those of Western Europe, are depicted presented in isolation. with halos in art. In Islamic art, halos are First, light streaming through a doorway dispensed democratically. They appear on the behind Christ surrounds his head, providing heads of ordinary human beings engaged in the radiance of a halo. Second, the doorway is good works, sometimes according to a pictorial surmounted by a curved molding shaped like a logic in which the artist seems to dispense them portion of a circle. If, following the tenets of at will. gestalt psychology, we imagine the circle Halos, like dragons, have literary antecedents completed, its location on the picture plane is that explain why they are depicted in painting that which would be occupied by a large disc though nobody has ever seen one. The prototype halo around Christ's head. Or his head is at the for the halo is sometimes identified as the center of the implied disc. flames that appear around the head of the infant Leonardo's third intimation is of a line or Ascanius in Virgil's Aeneid. Among older ray halo, created by locating Christ's head so images, fire is more directly linked to the that it overlaps the vanishing point of a potency of the godhead in the story of onepoint perspective system. The lines usually Semiramis, consumed by flames when Zeus said to be receding to that vanishing point grants her request to show himself in his true might, with imagined reversal of their form. In the Old Testament, when Moses comes imagined direction, be thought to proceed from down from speaking with God on Mount Sinai, Christ's head to form a world-encompassing beams or rays of light shine from his face, ray halo. Leonardo has played cleverly on that which he covers with a veil except when truism central to many optical illusions: all speaking with God. (Exodus 34:29-35). The lines point in two directions at once, if lines horns growing from the head of Michelangelo's "point." Moses brought attention to the passage. The The fourth intimation is psychological or Hebrew word karnot, it was pointed out, can metaphysical. The miraculous is essentially a mean either rays of light or horns, creating a condition in which more is sensed than can be possibility for mistranslation. seen. Christ's halo is transcendental (or mirac- If the theory is correct, which meaning ulous) in Leonardo's Last Supper because came first? Did karnot originally mean horns although it is not there in one sense, we are (a symbol of male sexual power), later adapted 120 Light as Symbol and Visual Metaphor II

to identify rays of light or rays radiating from a infants. In Renaissance paintings of the human head? Or were the circumstances the Annuciation, white lilies symbolize the purity of reverse? In English, names for light are not the Virgin Mary. usually derived from names of other objects. Black, color of death or of mourning, More often, names for light inspire names for expresses regret for lost loved ones. It also other objects. The word radius, as in radius of a recalls loss of sexual innocence, the event by circle, comes from the Latin word for ray. The which, according to St. Augustine's rays in halos are radii of a circle that has the interpretation of Genesis, death was brought human head at its center. The word "luminary," into the world. Hence the otherwise curious sometimes used to mean just an important association of black clothing with funerals but person, is more correctly a person who also with sexual provocativeness, as in women's illuminates the human race, a borrowing from black underwear or the idea that black dresses Latin fume, light. for women are sophisticated. Black or darkness The Moses of Exodus becomes an is associated with prostitution, as in the phrase independent radiator of light on Mount Sinai, "ladies of the night." Baudelaire, following the although with equal literary logic he might have symbolic cue, compared Parisian prostitutes to passively reflected the light of God. When the black cats. figuration is carried over to the visual arts, rays The Bible identifies carnality as the first sin of light rarely or never shine from faces. They and says that the wages of sin is death. Or so rest behind heads or around them, a variation the medieval moralists tell us, taking their cue that more easily lends itself to pictorial from St. Augustine.Black is understandably the representation. color of both death and sin, ambiguously both sin generally and sexual transgression. The association is not limited to Judeo-Christian Color Symbolism societies. Islamic legend says the sacred black By comparison with symbolism of light, color stone of Mecca was white when it fell from symbolism is diffuse. It probably developed at a heaven. The stone, incorporated today into the later date and exists largely as a supplement to masonry of the Ka'aba, became black-rather the symbolism of light. Symbolic associa- than blue or green-from the sins of the tions with darkness and light have international human beings who touched it. Why was the roots and appear in many . Associations stone sent? Evidently to remind human beings with the hues, which might have developed of the blackness of their deeds, for which they during the Middle Ages, seem more closely could expect to be punished. linked to the moral imperatives of the Old The symbolism of black and white rests Testament, although color names appear rarely solidly on religious metaphor, on its intimation in the Bible. that the greatest death is the death of the soul. Symbolic associations with black and The symbolism associated with other colors white, carried over from those of darkness and usually turns on association with objects light, are based almost entirely on the play of typified by a particular color. Blue is cool positive against negative. White is a symbol of because it is like the sea or because we imagine purity, of the original sexual purity of human ourselves turning blue with cold. But red, life. The symbolism drifts over to pale colors followed by orange, is closely linked to black in general, as in the traditional use of very pale and white, to their intimations of sexuality, sin, pinks, blues, and other tints that approach and death. white for the clothing of sexually innocent Red is passion because it is associated with Light as Symbol and Visual Metaphor II 121 the colors of blood and fire. The passion is sex- we need not assume that each member of each ual and gender differentiated. Red dresses, like group thought about them. Modern prisons are black dresses, were once thought too daring for said to have been inspired by the ideas of the presumably virginal young women. Sexually Quakers, who thought that incarcerating crimi- aggressive men are not thought to dress entirely nals so they could think about their sins was in red. But we might imagine them driving red preferable to hanging them. The traditional sports cars and being called redblooded. black-and-white stripes of prison clothing Women are not called red-blooded. As in the reminded the miscreant of the need to think title of Stephen Crane's Red Badge of Courage, about right and wrong. It also served as a red for men has positive connotations and reminder of the austerity of prison life, the means brave and manly. As in the scarlet separation from pleasures including the plea- woman of Revelations 17 or the title of sures of color in clothing. Gray, another popu- Hawthorne's The Scarlet Letter (the scarlet let- lar color for prison clothing, is similarly ter A was branded on an adulteress), red for austere. women suggests sexual transgression, associ- Among religious groups for whom the ations similar to those for black. world is a source of temptation, such as the A red flag means war, aggression. A white American Puritans, colors other than black and flag means surrender, including surrender to white are often avoided in clothing. The the will of God. A black flag means piracy, reasoning, presumably, is that black and white aggression against decent society. The para- have moral associations and other colors do digm for the moral associations of red, white, not. Other colors, therefore, serve no reli- and black is the Biblical story of the Garden of giously useful purpose and can be classified Eden, interpreted by medieval moralists to with jewelry and perfume as unnecessary mean that Eve and her daughters led men adornments of the body. astray. Restrictions on the colors of clothing worn Black, white, and the proverbial touch of by men are more common than those for red are a popular color combination, juxtapos- women, reflecting the traditional idea that the ing a bright hue against both extremes of man is a role model as head of the family. achromaticity. The taboo against using large Mod-ern Hasidic Jewish men in New York areas of red in home decoration, or excessively City wear black business suits and white shirts. bright varieties of the color, reiterates the In rural Pennsylvania Amish men wear black theme that red easily gets out of control. No suits or denim work clothes and bright parallel taboo exists against too much black, ultramarine blue shirts. Amish women wear although large amounts of black in home deco- blue, green, brown, or violet dresses, avoiding ration or in a young person's clothing might be red, orange, and yellow, the warm colors. considered depressing or macabre. In the Dresses are one color only, probably for the 1940s, red combined with pink, orange, or same reason snap fasteners are used instead of purple was regarded as a bad combination; the buttons. To cover the body with clothing is colors were said to clash. Pink, orange, and decent, but more than one color in shirts and purple can each be mixed using red and there- dresses is not needed. fore are red related. The issue, again, is a need Rules about allowable colors in clothing are to avoid too much red. not unique to religious communities. Business Moral associations need not dictate the suits in dark colors for modern men, a uniform worn every day. When they in many large corporations, bespeak dedica- do, the associations are easy to read, though tion to the job, a renouncing of all distractions 122 Light as Symbol and Visual Metaphor II

including bright or light colors in clothing. The clothing in early Renaissance paintings. The Hawaiian shirts in strong colors worn by the practical purpose served by colors other than president of an innovative computer software black and white is that when the pope or his company assert a right to other standards. cardinals appear in public, the colors of their People are sensitive to the messages conveyed clothing cause them to stand out from the by colors in clothing, intolerant if invisible lines crowd. are overstepped, and more critical of men's Red is thought of as an active color, hence clothing than women's clothing. masculine. Blue is calm or passive, hence femi- Despite a widespread elimination of dress nine. The associations can also be reversed. codes and virtually complete freedom in lei- Red has been characterized as emotional, thus sure wear, creative costumes in creative colors feminine; blue as intellectual, thus masculine. are not considered appropriate for people in For the clothing of infants, pink is for girls, pale authority or responsible positions. Hawaiian blue for boys. Blue suggests the vastness of shirts rarely appear during business hours on, blue seas and blue sky, leading to an associa- say, bank presidents or United States senators, tion with intellect and spirituality and to fas- even the senator from Hawaii. A police offi- cination with blue eyes. Red in its negative cer who wore an orange, purple, and green aspect suggests wantonness, shame, or infamy, shirt with his uniform would be ordered to consequences of failure to control the pas- change it. sions. Blue suggests melancholia, an inability Clothing not criticized for a high school to function. Rejected lovers in poems master student might annoy a judge if an attorney wore their blueness or die of broken hearts, a pointing it to court. If the judge felt the colors were too from blue melancholy to the blackness of the extreme to be consistent with the dignity of the grave. court, the attorney could be removed from the Purple suggests richness or royalty by courtroom. Many clients would not be pleased reminding of robes that ancient kings dyed to be represented by an attorney in pink with . The color is also associated overalls and a green shirt. The president of the with extreme rage, perhaps similarly a United States would be criticized if he wore a prerogative of kings. Or it suggests the overly vermilion jumpsuit to a summit conference ornate (purple prose). In societies where with the Russians. The Russians might be more elderly women were expected to dress in black, shocked than the Americans. We still equate pale shades of lavender, mauve, or other dignity with dark or subdued colors, an idea purples were often acceptable alternatives. with traditional roots. The Chinese are said to have developed a Freud pointed out in Totem and Taboo that distaste for milk because all milk products something prohibited on ordinary occasions is were once reserved for their Mongol con- often allowed under special circumstances. querors. The similar dedicating of purple to Priests and ministers often wear black suits or royalty may have achieved a parallel effect. robes with white collars, colors symbolizing Like orange, the color is not popular and was their dedication to God or renunciation of the rarely used in Europe in interior decoration or ways of the world. In the Catholic church, men's clothing. That purple is employed far cardinals wear scarlet robes; the pope can wear more extensively in Islamic art, often in com- all red or all white. The symbolic message is bination with red, added to its aura of the that higher echelon officers of the church wear exotic, an aura elsewhere reflected in the mod- bright colors for the glory of God, the same ern association of the color with homosexu- reason angels are shown wearing gorgeous ality, particularly male homosexuality. Light as Symbol and Visual Metaphor II 123

Associations with orange are as unfortunate name) appears to have positive connotations, as those with purple. The color is thought to while yellow has negative connotations. symbolize bad luck in Japan. In the West, The word yellow is avoided in English in orange and black are the colors of witches and metaphors that have a positive aspect. We have Halloween, rarely a popular combination for golden girls and fair-haired boys, but no yellow women's clothing. During the Middle Ages, people except cowards. People with yellow hair women with carrot-colored hair-redheaded are called blondes. Sunny days are golden, not women-were thought to be witches and yellow. Golden replaces yellow in many plant often burned at the stake. Later belief held that names, even if the flower or fruit is yellow: redheaded women were sexually promiscu- , golden seal, golden bantam corn, ous, another way of being a witch. The golden delicious apples. Silence is called golden association was picked up in Renaissance to show that we value it. Journalism is called painting. Mary Magdalene, the repentant lady yellow if of a type we do not value. In a thread of the night, is traditionally pictured with red separate from the goldyellow link, the negative (orange) hair, a painterly device for identifying association of yellow with cowardice must turn her and alluding to folk beliefs about hair of on attributing to the frightened a tendency to that color. involuntary urination. Why are people with carrot-colored hair Goethe, in a passage expurgated from East- called redheaded rather than orange-headed? lake's translation of Farbenlehre, found that, Whatever the answer, associations with orange "when a yellow color is communicated to dull hair follow those for red in being gender and coarse surfaces . . . the beautiful impres- specific and sexually allusive. Redheaded men sion of fire and gold is transformed into one were not thought to be witches or devils in not deserving the epithet foul; and the color of medieval folklore. Only women with hair of honor and joy reversed to that of ignominy and this color were characterized as enemies of aversion. To this impression the yellow hats of God (witches) or bewitchers of men who bankrupts and the yellow circles on the mantles might tempt them to promiscuous behavior. of Jews, may have owed their origin. Cuckold The paradigm, again, is Eve, and the yellow is really nothing but a dirty yellow" associations rest on a foundation of religious (Matthaei 1971, 169). metaphor. The practices Goethe reports can be related Green is associated with the verdancy of to money as easily as to ignominy. Bankrupts growing things. In its negative aspect, green is were disgraced because they mishandled linked with envy. Many molds and fungi are money (and therefore were made to wear green. Unembalmed corpses, if Caucasian, yellow hats). The color name cuckold yellow acquire a greenish hue after death. Envy, we (no equivalent exists in English) also implies are evidently to understand, rots, consumes, loss of wealth and disgrace because of this kills, or is deathly. mismanagement. The wealth in this case is of Yellow suggests radiance and joy because another kind. If a man's wife is regarded as his associated with sunniness, the yellow of the possession, her infidelity suggests the sun, or the richness of gold. The dense layer- cuckolded husband has allowed himself to lose ing of symbolic association for the color may what is rightfully his. be more complex, even, than that for red. In Jews, barred from many professions during German,Gold is gold, Geld is money, and gelb the Middle Ages, were permitted to be money- is yellow, a similarity of sound that suggests lenders, an occupation despised because all three are closely associated.Gold (the color interest was charged. Compelling Jews to sew 124 Light as Symbol and Visual Metaphor II

yellow circles on their mantles suggests, again, legacy we need not accept. Color and the vis- that the Christian majority associated this ual arts are visual, adequately understood only group with money, and in a negative manner. on that basis. Most people know whether they The reasoning was recycled in Nazi Germany, like Egyptian art or the art of India before they where Jews were compelled to wear yellow discover what the symbols in the works mean. identifying stars. The color was meant to be an Some never care what the symbols mean. insult. Those persons unable to decide what they like until all the symbolism is thoroughly explained, and until they know, say, the The Future of Color Symbolism pharaoh's name and his wife's name and what Freud said nothing about color symbolism but the Egyptians ate for breakfast, are not visually a great deal about symbolism generally that is oriented. They absorb information like sponges pertinent. He pointed out the symbolic content but only information of certain kinds. of dreams and fantasies, their hidden sexual I see no danger that color symbolism will content, the creation of symbols by attributing inspire pogroms against redheaded women, gender to what is not human (blue and the sea though taking symbolic ideas about black and are regarded as feminine), and the cloaking of white seriously contributes nothing to improv- ideas in symbols. His patients preferred not to ing race relations in the United States today. know what was on their own minds, to avoid The conceptual problem with color symbolism, feelings of guilt. They punished themselves by as with color theory, is its reliance on developing neurotic symptoms, acting out the simpleminded slogans. Thousands of variations need to atone without recognizing why this of, say, red exist. Developing visual sensibility need was felt. to color implies noticing the wide range of The proposed therapeutic solution was to variation. The English language unfortunately interpret the symbols, freeing the patients to encourages lumping all variations, no matter understand their thoughts and deal with them how different, under the label red. Most people directly. Although Freud recognized the tradi- follow the cues and do this. Color symbolism tional nature of symbolism and its appearance reinforces this oversimplification by reducing in folklore, he failed to follow this thread to its the effective aspect of color to nonsense conclusion. Sick patients grow up in a sick sayings. It deadens people's ability to notice society. Color symbolism and other symbolism their feelings. Does every variation of red make is traditional, passed from one generation to everyone think of passion at all times? Why the next. If symbolism in most cases confuses would this be the case, when reds can be so thought rather than refining it, as seems to be different one from the other? true of color symbolism, little can be accom- The moral thread running through color plished by psychoanalytic treatment of an symbolism is discouraging on other grounds. individual patient. Subliminal though it be, this morality I do not Freud was pessimistic about human possi- care to perpetuate. People over the ages read bilities. Jung saw more positive aspects in sym- the Bible selectively and remembered what bolism. Symbolic associations were treasures supported their biases. We preserve their mis- that drifted up from the depths of the collective takes by attributing significance to their sym- unconscious. They were the content of that bolic ideas. A conception of morality that unconscious. Symbolism, Freud and Jung regards wrong and right as absolute, compar- would have agreed, has a literary content. I ing them to night and day or black and white, regard it as excess baggage for that reason, a is insensitive. The moral issues facing the Light as Symbol and Visual Metaphor II 125 human race today have nothing to do with not accomplish. We can and should stop whether scarlet women wickedly seduce men psychologizing about these associations. We by wearing irresistible black lace underwear, have more refined ideas about morality than with whether redheaded women are witches color symbolism conveys. And we need a and whores. more sophisticated understanding of color We will continue to regard red sports cars as than can be reached by this route. Colors have sexy, and black lace underwear as devilish. no correlation with the sexual behavior of Eliminating traditional symbolic associations human beings or with human ideas about with color and light would require decon- morality. Why would anyone ever think one structing the English language, which we can- could shed light on the other? CHAPTER 15 Nonvisual Seeing and the Metameric

The Pythagoreans, having constructed matter out of numbers, next proceeded to arrange the main members of the universe according to a plan in which there was a little observation of nature, and a lot of a priori mathematical reasoning. Benjamin Farrington, Greek Science

olor is something we see, rather than Vermaasen said he could identify colors by something we hear. But the question their relative smoothness. Boyle suspected of whether color can be experienced sensitivity to the smell of fabric dyes, pointing by nonvlsual methods is ancient. The issue is out that Vermaasen"seems not consonant to whether human beings can transcend the himself about the Red, which as you have seen senses, possibly by the power of pure thought. in one place, he represents as somewhat more Classic anecdotes about remarkable gifts are Asperous than the Blew; and in another place, usually offered by narrators who heard the very Smooth." With the wistfulness of the story from someone else. Robert Boyle reports faithful who missed the miracle, Boyle longed about John Vermaasen, a thirty-three-year-old for "the Opportunity of Examining this Man my Dutchman blinded by smallpox at age two, self, and of Questioning him about divers who was reputed to be able to distinguish particulars which I do not find to have been yet colors by touch (Boyle [1664] 1964, 42-49). thought upon." The "Ingenious Person" who told Boyle about The question of whether human beings can the prodigy had tested Vermaasen by asking see without eyes is moot. Any method of see- him to sort black, white, red, blue, green, yel- ing that is nonvisual is, by definition, not low, and gray ribbons. The test was inconclu- vision. Furthermore, we cannot hear with our sive in that "he call'd the White Black and the noses or see with our ears. No sense organ is Red Blew." able to deliver percepts of the type specific to

126 Nonvisual Seeing and the Metameric Grays 127 any other. The dream of nonvisual seeing the Punic Wars is not germane to the question perennially recurs in the media and in journals of whether one was there. Blind scientists who of psychology, philosophy, and philosophy of feel as if they have seen red and therefore as science (Hattersley 1970, 55-58; Perkins 1971, if they understand the experience are in a no 329-37). In academic forums the question often more defensible position. Either they saw the takes the form of whether a hypothetical blind color or they did not. "As if" is not good scientist could learn as much about color from enough. instruments as others learn through using their The literature expresses greater interest in eyes. Can knowing about red equal or exceed whether the blind can experience color than in knowing what red looks like? Can so much be whether the deaf can experience sound. The known about an experience that having the argumentation is often embellished with experience adds nothing further? Can thought technological trimmings, as if they could replace all other modes of human activity? We provide the necessary . Feigl asked that are never asked to consider whether "knowing his readers imagine "the case of a congenitally about" color is possible without a brain. We blind scientist, equipped with modern electronic assume that the brain and its function are instruments who could establish the inseparable, an assumption not made about (behavioristic) psychology of vision for subjects eyes. endowed with eyesight. The blind scientist Knowing about something implies could thus confirm all sorts of statements about acquiring knowledge. Knowing what visual sensation and qualities-which in his something looks like implies undergoing knowledge would be represented by experience. Because knowledge is not hypothetical constructs" (Feigl 1958, 385). experience, the blind physicist is at a Feigl's blind scientist could not confirm all disadvantage. Although knowledge can be statements. He could evaluate those that acquired from experience, experience cannot required knowledge of color, but not those be extrapolated from knowledge. The requiring experience. His own experience limitation is not surmounted by redefining would be limited to the reading of instruments. experience to include empathy, which is said This is experience of a different order than the to give people insight into experiences they experience of those who see or who apprehend have not undergone. Empathizing with color by looking at it. Given the different another's experience, and thereby acquiring an experiences of the sighted and of the blind, two understanding of it, is a bona fide experience different types of knowledge can be derived. for the empathizer. But it differs from the This need not imply, except in a gross practical experience with which empathy is sought. sense, that Feigl's congenitally blind scientist is We cannot be more than we are or do more disadvantaged by comparison with those who than we do, a constraint that has an objective are sighted. He cannot understand the basis. The criterion for an experience, even an experience of seeing. But the sighted, as bias inner experience, is that it occurs. This effec- leads us to forget, are unable to comprehend the tively means it has a spatiotemporal location. experience of having never seen. If an experience is in the mind, its location is Feigl's blind scientist differs from a that of the body with which the mind is randomly selected blind person. The scientist associated.Arguing in favor of experiences has unimpeachable professional credentials. As that do not occur or that do not occur at loca- important, he has instruments at his disposal tions appropriate for those experiences is self- that are both "modern" and "electronic." The contradictory. Feeling as if one had fought in instrumentation, although it may include spec- 128 Nonvisual Seeing and the Metameric Grays

trophotometers and colorimeters, is as idealized Forty years ago, Evans found that "the (or as improbable) as the scientist. No difference in color that can be detected by an instrument presently in use, no matter how expert between two otherwise identical pieces modern and electronic, enables the blind to of paper is, at present, smaller than the solve even such simple color tasks as sorting difference which can be detected by red poker chips from green ones or spectrophotometric measurements and distinguishing red lights from green at traffic calculation" (Evans 1948, 203). No significant intersections. change has occurred since that date. The question about instruments for Feigl was not interested in how measuring color is not really whether they could spectrophotometers work or how well they work miracles for the blind. It is how reliably work. Like most of us, he took the wonders of they perform their intended tasks for anyone. modern machines and electronic Wright pointed out that in the checking and instrumentation for granted, though instruments calibrating of colorimeters, interlaboratory for measuring color fall far short of public comparisons "have shown a depressingly large expectation. What interested Feigl, and what is spread in colour measurements made with supposed to be proved by the imaginary blind instruments of the same design, iet alone with scientist, is the supremacy of understanding different types of instruments" (Wright 1962, 3; over experience. Feigl's purpose was to defend [19441 1969, 258). deductive thinking, including those deductive Kelly, who identified six methods for meas- techniques applied to the analysis of color that uring color, found that, for maximum have yielded so little of substance. We do not accuracy, "a color should be measured by a need more theories of color devised by those spectrophotometer and the result expressed who reason and write as if they had never seen numerically either in terms of the CIE method color. or a Munsell renotation" (Kelly, 1965, 2). Without prejudice to whether hypothetical Deane B. Judd of the National Bureau of Stan- blind scientists can become experts on color, dards did not find this was sufficient. Judd real scientists with bona fide visual handicaps arrived at a conference on colorimetry to find the going more difficult.The British chem- report he had been unable to complete an ist John Dalton (1766-1844) found he made experiment as scheduled. The data had been errors in identifying the colors of chemical collected with a Beckman spectrophotometer. solutions and precipitates. After comparing his When an attempt was made to repeat the meas- color observations with those made by others, urements using a General Electric spec- Dalton concluded his evaluation of red and trophotometer, problems arose from "dif- green was defective. Describing his condition, ferences in the spectrophotometric values for he wrote the first scientific report on color the samples obtained respectively by the Beck- blindness, for many years called Daltonism man and General Electric spectrophotometers. after him. Those with whom he discussed his With the values given by the latter the good vision must have tried to explain what red or agreement previously noted no longer held. green meant to them. It is not recorded that Some further work on the spectrophotometry this cured Dalton. He was precluded from was required" (Symposium on Visual Prob- acquiring a type of understanding that could lems of Color 1961, 343). The hazards of compensate for inability to experience. The switching brands are not to be ignored when kind of superior understanding Feigl imagines measuring with a spectrophotometer. Further- simply is not possible, whether for imaginary more, how refined is a spectrophotometer? blind scientists or for real scientists like Dalton. Nonvisual Seeing and the Metameric Grays 129

A less famous example of a scientist with a the scales on the instruments were in braille, visual problem is provided by J. Plateau, the who is to set up the instruments? No blind scientist who published a book on surface scientist can follow, unaided, a command to tensions in soap bubbles. In studies of surface select five blue objects to be studied by a tension in thin films, observations of color are spectrophotometer. important. Soap bubbles show bands of color on The issue of the hypothetical blind scientist their surface, as well as a black area. If the is linked to that of whether sense data can be black area is touched, the bubble bursts. But a explained in a manner that can support a sharp instrument can pierce other parts of the notational system. Like Aristotle's hypothetical surface and be withdrawn without breaking the blind man who argued about colors, bubble. James Clerk Maxwell, in the review congenitally blind scientists are precluded from "Plateau on Soap Bubbles," explains how this understandings similar to those of sighted scientist was able to continue his research after persons until this can be accomplished. Sense losing his eyesight (Maxwell [1890] 1965, data must be exhaustively understood and 2:393- 99). Like John Milton, who asked translated into abstract (nonsensual) terms in a members of his family to read to him, Plateau manner ensuring that nothing of consequence is borrowed the eyes of others. Friends and lost in the translation. assistants performed experiments under his The claim of contemporary physics is that direction and answered his questions about the task is already accomplished. Colors, or what had happened. those relatively few colors included on the The experiences of real-life scientists with electromagnetic wave scale, can be notated in visual impairments suggest that the hypothet- terms of the wavelengths of the light they ical blind scientist who masters all by the force reflect. But too much is lost in this system. of intellect, assisted by modern electronics, is Consider the metameric grays. Developed by just a technocratic conceit. Nobody is capable the Inter-Society Color Council, these grays are of perceiving what for him or her is visually indistinguishable under specified imperceivable. Even if machines were able to lighting conditions. Yet they have dissimilar think or experience, the perception of color by reflectance; therefore, they could never a machine would be a perception for that correctly be coded by the same notation in any machine. A blind person could no more annex notational system based on reflectances. the machine's experience than he or she could Metamerism is an everyday phenomenon, appropriate the experience of another living familiar to all who mix paints. To match the human being. Machines cannot live our lives color of a sample, we need not know the for us, even if it could be shown that machines colors mixed for the original sample. Two were capable of a superior quality of life. paints that match in color but are mixed from Unless a priori knowledge can equal or different constituents would have dissimilar exceed the scope of knowledge that is a reflectances, although whether this would be posteriors (I do not think that it can), no blind perceptible to human beings depends on the scientist can understand color in the same illuminating light. This is why colors that manner as a sighted person. A blind scientist match under one light may not match under arguably cannot confirm any statement about another. color, because verification is inseparable from The problem in explaining a metameric a need to confront colors, at some point, with relationship between two colors to the a functioning eye. Although the blind scientist hypothetical blind scientist is that the notation might take instrument readings competently if for the colors, which confirms dissimilarity in 130 Nonvisual Seeing and the Metarneric Grays

reflectance, would be contradicted by the ditions, we see no more difference between perceptual experience of seeing them-they them than would be seen by the scientist. would look alike. Under some lighting An invention as bizarre as the hypothetical conditions, a paint made from, say, inherently blind color scientist may not be needed to green pigment might exactly match another reach to the nature of the difference between paint made from a mixture of blue and yellow knowledge and experience. We all have vision pigments. The two paints would have different that is limited in a variety of ways. I cannot see, notations but would look alike in color. How for example, cosmic rays. I fail to understand could the blind scientist be edified on what how this limit could be transcended by learn- "look alike" means? The term is ing a great deal about cosmic rays, irrespective comprehensible only as a report of an of how recondite the acquired knowledge or observation by a functioning eye. how sophisticated the instruments used to Speakers of English can make two classes of collect it. Knowledge cannot aspire to the statements in which color names are used. Only complexity of experience and therefore cannot those statements that refer to language rules can substitute for it. be understood without looking at the colors, or To suggest that the blind will never see is without an ability to see colors. To say two foolhardy. Those without legs learn to walk. In colors look alike is to report a visual blindness caused solely by cataracts, vision can experience, a report understandable only to be restored in many cases. Among myopics and those who have had similar experiences. astigmatics, contact lenses and eyeglasses Blindness, however, can be considered as a improve otherwise substandard vision. But this relative condition. Is a person either blind or not is the realm of the prosthesis, the transplant, blind? The either-or categories reduce the and other wonders of modern medicine and diversity of experience to simpleminded optometry. Whether vision can be restored for stereotypes. People classified as legally blind the-profoundly blind depends on whether eye can see, though their vision cannot be corrected transplants are possible or a functioning to better than 20/400. Even the best lenses allow artificial eye can be constructed. These are them to see no more at twenty feet than the different questions from that of whether the statistically average observer sees at four need for a functioning natural or artificial eye hundred feet. Among those who are not can be bypassed through knowledge, color congenitally blind, entirely blind, or legally notational systems, electronic instruments, blind, many people need eyeglasses for various human will, or blind faith. visual corrections. Everyone is blind to some Whether sensory experience can be extent, in the sense that no human being has eliminated is less important than why anyone unlimited visual acuity. would want to prove it to be superfluous. For The metameric grays mark one kind of limit Planck, eliminating sense perceptions from for everyone, the only reasonable explanation physics represents a "sacrifice made to pure for why they look alike although demonstrably knowledge. "But knowledge is not necessarily different. A species with greater visual acuity more pure than sensory experience, nor is than our own might be able to see the thought necessarily more pure than perception. differences we can just measure. Our own It leads to results more likely to be regarded as blindness (what we cannot see, rather than invisible and therefore superior. what we can) is what we might describe to the What is at issue is a system of morality in blind scientist to explain the nature of the which thought is regarded as pure and sensory metameric grays. Under certain lighting con- experience as morally suspect. The hypothet- Nonvisual Seeing and the Metameric Grays 131 ical blind scientist who knows all about color eliminate sensory experience because its is presented to raise the question of whether nature is unique. Thought is different, and one sensory experience can be eliminated, whether cannot be a substitute for the other. it performs no useful function. We cannot

PART THREE

Color and Form

The Eye sees no forms. It only sees that which differentiates itself through light and dark or through color.

Johann Wolfgang von Goethe, Color Theory

CHAPTER 16 The Two-Dimensional World

I think of color as an interval of space-not as red or blue. People used to think of color and form as two things. I think of them as the same thing, so far as the language of painting is concerned. Stuart Davis, Conversation with James Johnson Sweeney

the question of what causes light and and light mean the same thing. They refer to a color, that staple of the physical single phenomenon burdened with more than sciences, is matched in philosophy one name. and the life sciences by that of whether color is real. The answer to the first question depends on what causality is understood to Illusion and Reality imply. The answer to the second turns on the Color exists in the visual field. This two- difference,if any,between an entity that really dimensional universe, separate from the three- exists and one that just appears as if it might. dimensional world of touch, operates according Color cannot be shown to be either more to rules of its own. The visual field is neither or less chimerical than anything else.Thus,the the world nor my entire perception of the question of whether color is real invites the world. Beyond the boundaries of visual counter question of whether anything is real. experience, I can detect tastes, sounds, and A less incendiary point is that color is smells. I can experience spaces by moving inseparable from both the mode of perceiving through them. I can experience objects by it and the domain in which it is encountered. touching their surfaces. Another class of per- Though this is not immediately obvious,color cepts tells me about myself: whether I feel hun- and vision mean the same thing, just as color gry, tired, disoriented, thoughtful. The visual

135 136 The Two-Dimensional World

field has no direct connection to these other chair correctly, theory says, the chair would perceptual universes. It contains only my look the same at all times. But the camera, apprehension of what the world looks like. which lacks human fallibility, sees the chair A real chair has not entered my eye when I essentially as I do. The similitude is not entire, see one, any more than a real chair has entered but the congruency is remarkably close. The my head when I think about one. The visual camera even sees, in its own way, the colors field contains images, not objects. The English that may not be real. Either my percepts are less language is poor in forms for describing these illusory than I have been taught to assume, images. Language in general encourages the or the camera mechanistically shares my expediency of summarizing experience, delusion. including visual experience, in generalized or idealized terms. Perception, unable to encompass this economy, stubbornly wends its The Real World own way. Language and acculturation cannot affect The conceptual imperative is unambiguous: people's eyes. But they affect how we interpret the environment, like every object in it, ought to what we see. What we are taught in this case is be envisioned as stable, each a receptacle for its quite odd. The multiple presentations of the singular identity. The world, although it imagined chair, or of any other object, are to be changes, remains the same world it was accounted for by envisioning, so to speak, an yesterday. A yellow chair by the window in my objective object, divorced from any singular room will not be a different chair if moved to experience of seeing it. another place tomorrow. The dynamics of The transcendental object-in this example, vision, however, are relative. In my perception the objective yellow chair-is assumed to exist of world or chair, the presumed singularity each time anyone encounters any of its cannot be confirmed. The way any chair looks is individual presentations. It still exists (and affected by my vantage point, my perspective, remains yellow) when no one is present to look or, more formally, my coordinate system. at it. This objective chair is conventionally said I notice a chair on Monday from far away; to be real and to exist in a world of similarly on Tuesday, I see it nearer. On Wednesday, I real objects. We are asked to prefer this story look at it from a seated position, although I had about what cannot be seen to what is visually previously been standing. On Saturday, self-evident. someone is sitting in the chair and lighting The real world containing the real chair, conditions have changed so the chair's color although regarded as three-dimensional, is not looks different. On Sunday, the chair has been an exact equivalent for the three-dimensional moved to another room. If a camera were to world available to touch. It resembles a matrix record these sightings, no two photographs of superior to the universe of any individual form the chair would be alike. They would show of sense perception. The objective world different views of the chair. Nelson Goodman includes both corporeality and colors (the used the term presentations for these multiple objective chair is yellow), although I cannot apprehensions of a singular object (Goodman touch chairs with my eyes or see colors with 1957, 127). my hands. Extrapolated from the sum of the Conventional wisdom holds that the mul- perceptions, the real or objective world exceeds tiple presentations are evidence of the delusion the limit of perception. Reality is a useful of the senses and that the colors of the chair fiction, as is objectivity. What is the fiction may be similarly illusory. If I were seeing the useful for? It is necessary? It is reasonable? The Two-Dimensional World 137

Does it further understanding of color? been confirmed by human beings. An earlier As if we had passed through Alice's looking chaos of unspecified form (a darkness on the glass, the objects located within the objective waters) is said to have preceded whatever form world possess unusual attributes. Unlike the of creation occurred. particular presentation of a yellow chair that I That the world is of earlier vintage than alone saw on Tuesday, the objective, real, human beings is scarcely doubted. Why do we yellow chair, in its objective, real world, can be believe this? Some of the faithful cite Genesis. seen under an infinite number of conditions by Those who prefer another authority point to the an infinite number of persons at an infinite geological and biological record presented by number of times and places. This identifying fossils, or to astronomy and cosmology: to look characteristic remains stubbornly theoretical. back time and the big bang. But both scientific Nobody understands how to implement it and religious explanations are elaborate forms except the Mad Hatter. No human being is on of conjecture, neither supported by anyone's record with the heroic claim that he or she immediate experience. No human eye saw the personally observed any given object from world before human beings were on earth. every possible spatiotemporal location. The Because no one saw, no one knows, and we task might reasonably be accomplished, given simply make inferences. We hope that our rules all the time in the world. No human being is for making these inferences are reasonable and allowed that much time, a limit we must not that we will remember the difference between ignore. inference and observation, between believing However ardently I may hope that the real and seeing. chair exists when I am not looking at it, its pre- Seeing is believing. But we believe in many sumed existence can only be known through things nobody has ever seen, including the belief, as an article of faith. Knowledge that significance of our ideas. Objectivity (as in "the the chair is still there when I am not looking at objective world") and reality ("the real it, unlike knowledge that the chair is yellow, is world") are synonymous for most purposes. never available through immediate perceptual Either conception relies heavily on tactile experience. The profound barrier, as children elements difficult to reconcile with the visual discover, is that I cannot see whether the chair matrix. is still there if I am not looking at it. Reality can be loosely defined as all the real Endeavoring to circumvent this familiar visual objects in the world, along with the empty (but limitation, I can solicit testimony confirming real) spaces between these objects. Each of the the existence of the chair from someone who objects is said to have a front and back at the observed it while I was not looking. This does same time, though we cannot see both at once. not speak to the question of what happened to The inconsistency is not genuinely resolved the chair when nobody was looking at it. with mirrors. I can no more prove that the Solipsism is feared because human beings world and its objects have the structure need to believe that the world extends beyond attributed to them than I can prove that human experience of it. Creation myths from absolutes exist. all parts of the globe, including the opening If what is known (or believed to be true) chapters of Genesis, throw down the gauntlet exceeds what any of us can see, this substan- with assertions about what is metaperceptual. tiates the familiar claim that differences exist With few exceptions, they declare that the between knowing and seeing. A dog is able to world was created before human beings see a yellow chair, with less or greater visual inhabited it, a condition that could not have acuity than a human. The dog is not expected 138 The Two-Dimensional World

to have worked though the philosophical verify that the chair is there when I am not intricacies of whether the chair, or its yellow looking at it. Sense data disappoints color, ought to be regarded as real. Does it expectations, but the expectations have not matter that I have worked them through? What been sufficiently scrutinized. benefit can I expect to derive after I have To explain what the word chair means is determined, once and for all, whether the world reasonably simple. To illustrate the scope of the is real? Is that question significant? What does term, a carpenter can be invited to construct it mean? some chairs and show how chairs ought to be used. This cannot address the question of what we mean by saying these chairs are "real," Real and Unreal Chairs "exist," or redundantly even "really exist." The Does the world go away when I close my eyes? realness attributed to them implies the essential The visual answer to the question is that I that the chairs are still there when I am not cannot very well tell with my eyes closed. That looking at them. Human value judgment is such human beings have traditionally demanded that this is held to be more significant than the more of an answer than that has been, I think, a oddity that I am still here (or suspect that I am) misuse of human intelligence. This misuse when not actively engaged in the act of looking fosters reasoning based on the metaphor of an at chairs. invisible God, in which what cannot be seen is For the philosopher F. H. Bradley, in assumed to be more important than what we Appearance and Reality, "I can not transcend see. Conjecture loses touch with observation, experience, and experience is my experience. and ability to reason about observation From this it follows that nothing beyond myself weakens. The mind is valued more than the exists; for what is experience is its &the self'se senses, a weakness with a direct bearing on states." In assertions about what exists and what theories about color. does not, the word exists, as in Bradley's Whether a real, objective world still exists sentence, is almost impossible to define. when I close my eyes is not a meaningful Bradley could as reasonably have concluded question. Like the inquiry about how many that nothing beyond himself is himself, a angels can stand on the head of a pin, it asks for tautology. a determination that cannot be made. Worse, Taking Bradley's solipsistic world at face the impression of a significant issue is created value, objects in this world disappear when not by a semantic trick, a play on words. being perceived; only the self can be regarded Thereafter, the syllogism goes from bad to as real. In another kind of world, the worse-if I cannot look at the objective world inhabitants might disappear when not engaged while I am not looking at it, something is in perceiving objects. A universe of either of wrong with my eyes. Possibly, then, the colors these two types (or of many other conceivable I see are not real, a nonsense issue that has types) would look exactly like the universe in persisted for centuries. which we reside. To determine its actual nature Traditional arguments against the credibil- would be beyond human capability. I cannot ity of sense data turn on the distinction know that unperceived chairs are there or that between knowing (or believing) and seeing. they are not there. I can be certain only that I Sense data is thought to be philosophically dis- cannot see them when I am not looking at appointing because it fails to confirm what is them, a visual limit to what I can know. believed to be true. Because I cannot look at My hypothetical yellow chair with its a chair I am not looking at, I cannot directly green, red, and orange cushion is convention- The Two-Dimensional World 139 ally called a real object, even though, in the concatenation of words means the same as the present example, it is only a figment of my other. One is a description of what lies within imagination. Some people feel strongly about the realm of individual and aggregate human maintaining a proper sense of reality. If I said I perception. The other is a brief summary of that had decided to imagine an unreal chair, they data, no more or less meaningful than the would insist this is impossible and that I must original data. Confusion enters when "the chair" be imagining a real chair or a real hypothetical is inflated to more than it can reasonably be: chair-as opposed to an unreal hypothetical when the chair is called real and the sense data chair. Convenient fiction or not, people take unreal, as if one could be severed from the reality seriously. other. The argument that chairs are real but The real or objective chair should not be perhaps their colors are not is familiar. Its confused with the Platonic ideal chair, that sta- problematic aspect is not the unrealness ple of undergraduate courses in philosophy. attributed to the colors but the realness Nor is it the generic chair that floats in the attributed to the chairs. The physical sciences, mind as a ghostly surrogate for chairs in carrying the quest for an ultimate reality one general. It is, however, as conceptualized and step further, have not contributed to clarity. artificial as either. Both the ideal chair and the We are asked to believe that even the chairs generic chair consist of those parts common to are not real, though their electrons and other all chairs. These parts may be elevated, in the subatomic constituents are. If Newton's case of the ideal chair, to a level of chair absolute time and absolute space have been perfection. discarded in the physical sciences, no The notion becomes untenable if we foundation exists for the conception of an stubbornly plod on to compile a list of what absolute reality: an objective world in which those parts ought to be. Color, as often the case, real chairs (or real electrons) are frozen is at the center of the inconsistencies. The ideal monoliths, forever unchanging. chair and the generic chair must each be some Any chair said to be real can be exhaus- color (the ideal chair is probably an ideal color), tively explained as a generalized or idealized because all chairs are colored. But neither chair chair that cannot be experienced in its totality. can be any particular color. No particular color No matter how often I look at any chair, I is common to all chairs. Nor is any particular cannot encompass the visual experience color the ideal color or an ideal color for a another person might have when looking at it. chair. I can not see the chair's back while regarding To see or visualize a color that is no its front. Nor am I able to see it from every particular color (or that is an ideal color) is conceivable spatiotemporal perspective and impossible. Therefore, we cannot imagine what under every possible lighting condition. either of these idealized chairs might look like. To refer to "the chair" is a verbal economy, This is of no practical consequence because we necessary because English is not well adapted are said to be unable to see them. The reason we to dealing with visual concepts. Using the cannot see or photograph them is related to their phrase "the chair," although inexact, requires incomprehensible colors. fewer words than saying "the total of the var- The real or objective chair, unlike its generic ious ways in which the chair has thus far and ideal cousins, is neither an abstract appeared to me, and to others, on the multi- summary of the parts common to all chairs nor ple occasions on which it, or one or another a refinement of the symmetry of those parts. of its presentations, has been observed." One We talk about it as if it were a single chair with 140 The Two-Dimensional World

the potential to be seen from all possible per- ments. The game is to determine whether sense spectives. The closest analogue for its parts data, including vision and color, ought to be would be to imagine a camera taking photo- taken seriously. The winner is the individual graphs of one chair from every available view- who is most adept at manipulating the word real point, every possible distance, and under in support of his or her views. every conceivable lighting condition. The pho- Much confusion about color and visual tographs would be infinite in number. Any of experience could be avoided if we more often them might represent the chair as some par- remembered that the conventional figure of a ticular person had seen it as some particular monolithic real chair with monolithic real time. But no individual's experience would colors is a hypothetical construct. We imagined include all the presentations in the photo- it and never saw it. Because of its graphs. allencompassing nature, the real chair can never Although the generic chair and the ideal be directly experienced in its entirety: no one chair are respectively no particular color and can see a single chair in "every" possible way. the ideal color, the real chair can be every More at home in the mind than among the color. Certainly a real chair can be painted any senses, the real chair is an ideal, a useful fiction. color and still remain the same chair. A chair But if it were called an ideal chair, people that can theoretically be every color (or any would confuse it with the Platonic chair, color) is as improbable as the chairs of ideal another type of ideal. color or no particular color. The image of a In assessing visual experience, we imagine chair that I see, that can be photographed, and the real chair situated in a universe of similarly that lies within the realm of the senses, can only real objects to which we attribute responsibility be (and must be) some color or some collection for our perceptions. An analogue for what I see of colors. at is a snapshot taken from within Absurdity is a good criterion for whether a that imagined world, a nation with inexact statement is a mental construction rather than a boundaries. We continue to wonder, on report of, or inference from, sense data. The occasion, how real objects differ from objects, real chair, always there, always the same, is as real chairs from chairs, real truths from truths. inaccessible to the senses as Plato's ideal chair. A definition of real is needed. But the word This objective chair is something we all know, eludes definition except as the opposite of or have entered into a societal conspiracy to unreal. sustain. But it is not verifiable. Nobody can see Questions about what is real, including all it, or see it in all it aspects. It is not a visual variations of the questions about color's reality, concept. are questions about how to use language. The camera, because it sees without Regrettably, they are often taken seriously as knowing, reports, presumably correctly, what is meaningful or even profound questions about available to be seen: a chair looks different in the phenomenological world. Operationally, both shape and color from different when an entity is said to be real, a value perspectives. Anyone unreconciled to this judgment is implied between it and another sensual phenomenon-in the final analysis, a item judged to be unreal, therefore of inferior basic phenomenon of color and vision-is quick value. In this good-bad classification system, to add the qualification "But it is really the same real is always better than unreal. Real can even chair." I reply, "But it really looks different." mean wonderful, which is why we have such My respondent and I are in engaged in word idioms as "a real human being." play, a good-natured jousting with value judg- The real world, an object of knowledge, The Two-Dimensional World 141 may be contrasted with the unreal (illusory) are not real chairs is that they are images rather world of visual images and their chimerical than chairs. Furthermore, real chairs are real colors. Imagine that we swept away judgmen- only by stipulation, because no physical tal code words (real, unreal) and demanded a difference exists between a chair and a real rational explanation for why our conception of chair. I prefer regarding the chair and its image a monolithic reality is more significant than as different objects in different universes, each visual experience. A convincing argument with its unique set of characteristics. One, to would not be easy to supply. The immediate keep peace with those who insist on the facts are only that the visual world floating in designation, is a real (three-dimensional) chair. front of our faces differs from the real world The other is an equally real (two-dimensional) we imagine. We see a visual universe of image. observable changes, not conjectured stability, Whether three-dimensional universes with manifested as a succession of different images their objects are real in a manner that two- that may be apprehended by me or by a cam- dimensional worlds with their images are not era. A discussion of the various presentations of leads again to the question of whether touch is the chair that I see implies a different uni- more reliable than vision. Traditionally peo- verse of discourse than discussing what I ple must have believed this to be the case. Our believe I know about the chair. conception of reality is founded on the idea that Anyone asked to show what is meant by a we confirm whether objects are there by real chair might reasonably point to a three- touching them. But I wonder why the ques- dimensional chair as its closest approximation. tion arose, as if the senses had to be graded and Because every real object has an unreal twin, one found better than the other. We need both the next endeavor should be that of locating an touch and vision, and they usually work in unreal chair. Everyone knows where to look. harmony. What is said to be "unreal," in this example, is Possibly mirages led early people to believe the two-dimensional image of a chair available that the eye is not always reliable, that visual to visual perception. images can appear without the tactile A two-dimensional image of a chair is correspondence expected in that case. Do regarded as unreal because it deviates from the objects exist that we can bump into yet not see? three-dimensional chair (the real chair) The black holes that modern cosmologists assumed to be normative. The word image, as imagine in outer space meet those requirements. in image of a chair, is sufficient to suggest an But no black hole has been found so far. Only ephemera: an apparition of something else that in stories about the supernatural do people is the genuine article. By extension, the entire collide with objects that cannot be seen. content of the visual field is often called a Regardless of what inspired the idea that the visual illusion, because what I see is not the eye is unreliable, the reasoning is defective. world. Instead, it is merely an image, which Because of the greater scope of the eye and its has two spatial dimensions and cannot be ability to perceive colors, we see many things shown to have three. If I close my eyes, the that have no tactile equivalent, ranging from image of the world goes away. But I am told rainbows to the horizon to words printed in that the real world does not. books. The idea that three-dimensionality is The image of a chair is held to be unreal real and two-dimensionality is illusory leads to not because we have measured it against some a conceptual morass. It obliges us to explain thing called a real image and found that they how height and width, illusory when they did not match. The reason that images of chairs appear together ( as in two-dimensional uni- 142 The Two-Dimensional World

verses), can become nonillusory (real) when reached directly by vision, at least not by vision linked with depth. alone. We can only look into, but never enter or Because the distinction between real and touch, that parallel universe of two dimensions unreal is hierarchical, the central issue is value suspended before our faces. judgment. The rhetorical forms built into lan- The two-dimensional world is devoid of guage encourage us to affirm or assume that depth. But it includes color, the element three-dimensional universes are significant in suggesting that two- and three-dimensional ways that those of two dimensions are not. An worlds are discrete. One is not part of the other, entity might be described as "only an image"; any more than the eye is part of the hand. rarely is anything called "only an object." The Despite the assumptions made in projective syntactical bias toward trivialization of what geometry, I see no possibility that one can be is not three-dimensional probably can be regarded as a projection of the other, except in a traced to the accident that, as three- nominal sense. No element accessible to touch dimensional bodies, we live in the three- on the surface of a chair-no bump, hollow, or dimensional world. The tripartite nature of other feature I can feel in its geography-might dimension in our environment can be affirmed reasonably be said to produce color if projected by touch, even with closed eyes. It cannot be on a two-dimensional surface. CHAPTER 11 Delusion and the Geometry of Visual Space

When we use the words "white color," we must not suppose that what we mean by these words is either something outside the eyes or something in the eyes. We must not suppose it to be any place at all. Plato, The Theaetetus The flowers have no colors; they send off physical vibrations, called vibrations of ether, but colors exist only when there are the mind and eye to transform these vibrations. E. W. Scripture, Thinking, Feeling, Doing

o ask whether what we see, including The strange thought that the phenomeno- color, is real or a delusion of the logical world can be graded in discernibly senses is to entertain a curious assump- different degrees of realness is traceable to the tion: that some aspects of the phenomenology- pre-Socratic philosophers. For Heraclitus (ca. ical world are more illusory than others. The 536-470 B.C.), the only constants are the vari- proverbial delusion of the senses avoids solip- ables of motion and change. His universe, like sism, a belief that the world disappears when that of contemporary quantum theory, was in a not looked at. A more radical possibility is condition of continual metamorphosis. Nothing introduced. Perhaps the real portion-only ever stood still. For Parmenides (470 B.C.), the this portion-of the world remains. The unreal world was stable and unmoving. Motion was a or illusory portion, which includes color, delusion of the senses. The concept proved to be removes itself to some other place. The name- enduringly popular, though we need not assume less limbo to which it retires is as distant as the Parmenides invented it. Phenomena that we do limbo containing chairs I am unable to see not know how to systematize, particularly visual because I am not looking at them. Can any- phenomena, are still called delusions of the thing be partly real and partly unreal, or more senses. real in some parts than others? Parmenides, whose frozen world is other-

143 144 Delusion and the Geometry of Visual Space

wise absurd, must have meant to convey a The film in motion argues for the sense of realities other than those that can be illusoriness of the motion "that is not there" to seen. The hidden reality is a world that eter- no greater extent than it argues for the nally endures, although the world discovered illusoriness of the forms we see in motion, through visual observation is continually which are as profoundly not there. Zeno passing away. Whether one world is more real correctly predicted how motion would look than the other has been reduced, in Western when stopped by a movie camera. But motion thought, to the issue of whether the world is stopped is no longer motion, although the made of forms (corporealities) or events. Is condition has no separate name. touch, by which we apprehend forms, more reliable than vision, which reveals only the endless permutations of visual space, time, and Hallucinations and Visions color? The issue surfaces in the classical Zeno assumed that what cannot be touched is question of whether light is a stream of particles illusory, even if it can be seen. The assumption, or a wave. Particles, at least as the Greeks which Ivins regards as central to Greek thought, imagined them, can be caught and held in the opens the door to an unmanageable hierarchy of hand. Waves cannot be arrested without ceasing relative degrees of reality. No reliable criteria to be waves. are available for sorting the real from the more In one of his two paradoxes that speak to the real, the most real, the really real. Whether the question of whether motion exists, Zeno (ca. physical world is real is a semantic question. No 490-430 B.C.) asks us to consider the flight of discernible difference exists between an an arrow. The arrow cannot be moving if just ordinary world and a world dignified by calling resting at an infinite number of points in space. it real. It seems to move, which led Zeno to a Because the world is one piece, we can conclusion similar to that of Parmenides. build consistent models by saying that Motion is a delusion of the senses. Or it is a everything is real or, as in Buddhism, by saying greater delusion of the senses than the that everything is unreal. We can consistently substantiality of the forms that move. The arrow regard the universe as paradoxical, entirely real was there but not moving. and entirely unreal at the same time. Or models As if cinematography had been predicted by can be built in which the issue of reality is Zeno, any reel of motion picture film of an avoided. We can call reality just a semantic arrow in flight consists of a sequence of issue, not meaningful. Confusion ensues when individual still pictures, called frames. In each we follow Parmenides in imagining that the frame, the arrow appears in a slightly different world can be partly real and partly unreal. position. The question raised by Zeno's paradox What glue holds the real and unreal parts is whether it is a delusion of the senses when the together? Why did Parmenides not admit the images appear to move in a motion picture flaw in his theory? He could not explain being projected. The answer I prefer is that we motion, so he dismissed it by saying it was not ought to know better than Zeno. It is a matter of there. common knowledge that any reel of film Parmenides was not laughed out of town for consists of a linked sequence of still what could have been viewed as evasive photographs. The sequence of pictures exhibits double-talk. The Greeks may have already movement when projected, a visual fact that can been partial to the idea that the world is a be explained in terms of the limits of acuity of mixture of illusion and reality, of things that the human eye. exist and things that just seem to. The real and Delusion and the Geometry of Visual Space 145 unreal parts remain stuck together in some such things, under circumstances that give cases but float apart in others. The reasoning cause for concern. The rest of us do not really is by analogy with touch, untempered by what see the hallucinated objects and events. we learn from vision. Touch teaches that we The individual who reports seeing things bump into objects but cannot bump into that cannot be explained in terms of our space. Space, consequently, does not really understanding of the external world elicits a exist, a muddled reasoning that still haunts us diversity of responses. These responses depend today. on what was seen and how listeners react to the From hallucinations to colors, many visual story. People who believe Moses saw God on experiences explained in terms of illusion or Mount Sinai may not believe I saw God at the delusion can be more coherently understood supermarket. Beyond the subjective without those concepts. Consider how the interpretations made about hallucinations by following events might be assessed, in terms of persons other than those who have them, these illusoriness or absence of it. experiences remain private affairs. If William Blake said he saw an angel in a tree, we hazard • I see a chair. more in doubting the vision than in doubting • I see a hologram of a chair. that angels exist. No person is privy to another's • I see a chair. But when I report this, others vision-or visions. We cannot prove that Blake say I must be hallucinating because they saw a tree, and we cannot disprove that he saw cannot see it. an angel. • I deliberately lie, saying I see a chair when Facts provide no authority in this case, I do not. because the facts needed are not accessible. The fact that the world contains trees does not speak Wrong conclusions can be drawn about to the question of whether Blake saw the tree what we see, as about anything else. But a per- the angel was sitting in. The fact that the world ception either occurs or does not. The limit contains no angels does not disprove that Blake leaves no room for the concept of a wrong or saw one. It only indicates that the cause of any mistaken perception: for thinking a chair was perceived image was not, so to speak, a real seen although this was not the case. In each flesh-and-blood angel who happened to be example except the fourth, I must have seen passing by. the image of a chair, though differing state- If I believe Blake saw an angel in a tree, I ments can be made about the circumstances will call his experience a vision. If I disbelieve, under which the experience occurred. Any I will characterize it as a hallucination. To error lies not in the observation, but in assume that the tree was seen but the angel was supposing that a corporeal, three-dimensional not is the height of presumption. We cannot edit chair is necessarily present whenever anyone what other people see or even what we see. sees one. Although I am able to change my mind, I Quasars, holograms, hallucinations, movie cannot alter my visual experiences. films, still photographs-all give evidence that visual images can be seen in the absence of the three-dimensional objects nominally said Optical Illusions to account for the images. Individuals are not Hallucinations and visions are private affairs, committed for psychiatric observation because accessible through reports that are secondhand they do not really see the visual hallucinations for persons other than those who experience they report. Their problem is that they do see the hallucinations and visions. Another class 146 Delusion and the Geometry of Visual Space

of visual field phenomena, the optical illusion, We expect, for example, that two lines of shares with the hallucination the characteristic equal length will always look the same length. of evoking metaphysical doubt. It shares with But, in the Miiller-Lyer arrow illusion, they do color the interesting characteristic of being not (see figure 17-1). We expect that when two subject to verification. More than one person patterns are superimposed, the third pattern they can see the illusions. will create can be predicted. But moire patterns The psychologist R. L. Gregory classes are unpredictable. We expect that straight lines some optical illusions as "figures which dis- will look straight without regard to the field turb" the eye, while others are figures that "are surrounding them. But, in the Hering illusion, seen distorted" (Gregory 1966, 133-36). The straight lines superimposed on radial patterns unusual aspect of the disturbing or distorting look curved. is its universality. A large number of people, What is deceived, in these and innumerable viewing an optical illusion, will each provide other instances, is the predicting mind rather roughly the same report about how the figure than the perceiving eye. The hidden assumption appears. As a result, no criterion can be enun- is that an object ought to look a certain way. It ciated for how optical illusions would look to ought to display qualia consistent with an observer who saw the figures without being conventional expectations about that object. confounded. To be confused by optical illu- Straight lines ought to look straight under all sions appears normal, while to remain uncon- conditions, if for no other reason than that they fused is abnormal. Just as no method exists for are straight. teaching anyone to see red as blue, no method This type of noncontextual reasoning exists for training an observer to see an opti- would mean that a ship on a distant horizon cal illusion in an undistorted manner, irrespec- should not look like a dot, because everyone tive of what Gregory might have meant by that knows ships are not dots. The unreasonable term. and unreasoning expectation is that the visual If everything we see-the "visual illusion" field ought to correspond to our model of an -should be regarded as phantasmagorical, the optical illusion superimposes a secondary layer of illusoriness that is troubling. If the world is not really the way it looks, what can we mean by singling out the optical illusion as a figure that is not really the way it looks? Is everything unreal and a few special things more unreal? Goethe had a simpler idea: only optical truths exist. I believe Goethe is correct, and distorted seeing is a meaningless concept. The defining characteristic of optical illusions is not the illusoriness attributed to them. Each illusion consists of a visual or graphic form that contradicts commonplace intellectual expectations. Most of the expectations are Figure 17-1. The Muller-Lyer arrow guided by our strange conception of reality, the illusion. The central line segment is the same length in each figure. It appears suprareality in which everything remains the longer in the figure with outgoing fins, same although this is not what we see. which is the taller figure. Delusion and the Geometry of Visual Space 147 objective world with a singular identity. We is the kind of `theory,' alas, all too prevalent in should see what we want to see. Furthermore, psychology-it is no more than a rather we ought to immediately discover for any misleading statement of what we wish to deviance an explanation that commends itself explain" (Gregory 1966, 142). as conventionally acceptable. Whether or not 4. The empathy theory "is that the observer acceptable, the more reliable explanation typi- identifies himself with parts of the figure . . . cally lies exposed in the putative illusion. As and that he becomes emotionally involved so the Hering figure demonstrates, straight lines that his vision is distorted rather as emotion need not always look straight. Optical illusions might distort an intellectual judgment" teach visual lessons we do not want to learn. (Gregory 1966, 144). Optical illusions are usually investigated by 5. The pregnance or good-figure theory, seeking a real cause for the putatively unreal suggested by gestalt psychologists, proposes illusion.The aim is to fit the figures into a con- that illusions "are supposed to be due to ventional understanding of phenomenological pregnance exaggerating the distance of features reality by finding a way to explain them that seeming to stand apart, and reducing the will not disturb that understanding, that will distance of features which seem to stand not make waves. This leads to psychologizing together" (Gregory 1966, 145; Kohler 1947, about an eye that sees incorrectly and a brain 99). that corrects the eye's errors. The literature is 6. The perspective theory proposes that "the thin on consideration of the other possibility: illusion figures suggest depth by perspective," that the eye is seeing correctly but the brain is an undefined term in this case (Gregory 1966, not thinking very clearly. 145). Each of these theories makes proposals about the workings of eye and mind but begins The Muller-Lyer Arrows from an excessively schematic description of Among optical illusions, the Muller-Lyer the figure said to be catalyzing those workings. arrow illusion is a frequent object of study To Gregory, the arrow illusion "is simply a pair (figure 17-1). Gregory has compiled a list of of arrows whose shafts are of equal length, one six theories about why subjects take the line in having outgoing and the other ingoing arrow- the left figure to be longer than that in the heads at each end. The one with the outgoing right, although both are equal in length. heads looks longer although they are both in 1. The eye movement theory "supposes that fact the same length" (Gregory 1966, 140). the features giving the illusion make the eye More to the point, the line that looks longer look in the `wrong' places," a reasoning occurs in the figure that is longer: the figure further explored by Jean-Paul Sartre (Gregory with the outgoing arrow heads is taller than 1966, 141; Sartre n.d., 46). that with the ingoing heads. The Muller-Lyer 2. The limited acuity theory proposes that arrow illusion is not an error of vision. It is an "we should expect the figure with the outgo- error in the estimation of length. Any investi- ing fins to look too long, and the one with the gation of it ought to seek an understanding of ingoing fins to look too short if the acuity of the rules followed in estimating size. That these the eye were so low that the corner could not rules are not efficient is evident. Human beings be clearly seen" (Gregory 1966, 141). saw the sun, moon, and stars for millennia 3. The confusion theory "suggests that cer- without correctly estimating the sizes or dis- tain shapes `confuse' the perceptual system. It tances of these objects. 148 Delusion and the Geometry of Visual Space

Double Entendre as Illusion The ability to appear two different ways is Optical illusions other than the Muller-Lyer not unique to the cube illusion, and it invaria- arrows follow a similar pattern of defeating bly indicates two different interpretations are commonplace expectations. Many involve involved. I have ridden the New York subway ambiguous figures, which can be read or inter- when I was in a bad state of mind. Other peo- preted in more than one way. These illusions ple in the car looked menacing and dangerous. are less illusory than they are parallels to ver- When I was able to quiet my mind, the peo- bal puns, in which a single word is suscepti- ple looked ordinary and decent. The people ble to multiple interpretations. Some illusions in the car did not change. My eyes did not in this class play on the conventions of per- change. My interpretation changed. spective. Examples include the cubes, ambig- In the cube illusion, we assume the eye is uously both concave and convex, that appear seeing incorrectly in noticing that lines on a to "jump" from one state to the other (figure planar surface are susceptible to more than one 17-2). reading. We do not assume the ear is hearing Which end of the lines forming a contour incorrectly in catching homophones, words for the sides of the cubes ought to be regarded similar in sound and perhaps also in spelling as farther away? Alternating the reading, which that nonetheless have different meanings. If a the viewer can easily learn to do at will, tog- possibility for multiple readings exists, notic- tm gles the cube's state between the illusions of ing that possibility can scarcely be character- concavity and convexity. The fascinating ized as an error. limit-- a bona fide visual limit -- is that the Verbal puns and their mechanisms have cubes cannot be seen as simultaneously con- been more thoroughly investigated than the cave and convex. visual double entendre often seen in optical illusions. In Jokes and Their Relation to the Unconscious (1905), Freud explored the the- ory that jokes and puns violate psychosexual taboos. Puns more immediately violate lan- guage taboos. They confound expectations about how words should be chosen and aggregated into sentences. The telling evidence that play on words defeats expectation (the expectation that a given word in a given sentence shall have a sin- gle meaning) is the degree to which word play annoys some listeners. They lack the sense of humor to feel at ease in a world in which objects stubbornly refuse to behave as they ought to. Spoken syllables can refuse to com- ply with an expectation that they point to a sin- gle word with one meaning. A picture of a concave cube can look exactly like a picture of a convex cube, no matter how strongly we Figure 17-2. Cube illusion. A line drawing of a con- feel that it ought to be one or the other. vex cube is geometrically congruent with a line What expectations are involved in the cube drawing of a concave cube. illusion? First, we expect three - dimensional Delusion and the Geometry of Visual Space 149 space to behave like a rigid grid in which near object and image, between a three-dimensional and far never exchange positions. If a camel cube and a picture of a cube. But images are is farther from me than a penguin and I do not not tied to objects by invisible strings. Pictorial change my position, space cannot "flip" so images are marks on a surface, entities in their that the penguin is farther than the camel. Sec- own right and not surrogates for the objects ond, we expect pictures on two-dimensional represented. Each of the several forms of the surfaces to behave as if they were counterfeits cube illusion is a diagram on a planar surface. of the objects they depict. If object a has cer- Each has the attributes of a diagram, not the tain qualities, we expect a picture of object a attributes of a cube. to show these qualities. Finally, we expect that Some people resist the idea that, say, a pic- if object a is not object b, a picture of object ture of a distant mountain in a Renaissance a is not a picture of object b. painting is just spots of paint on a canvas. They These expectations seem to have the force want to call the image an illusion of a moun- of the self-evident. Yet an inadequate concep- tain and theorize about what the term means. tion of the phenomenological world supports I conclude they want a label for the paint spots them, a conception that fails to properly dis- that includes the word mountain. This creates criminate between three-dimensionality and a literary link between the paint spots and their two-dimensionality.When dealing with images conception of the actual mountain, even if no on planar surfaces, say, a picture of a cube, actual mountain ever existed and the artist each of the three expectations is either dead imagined the scene. These people know how wrong or irrelevant. To deny this by insisting, to think about mountains but not about colors instead, that the picture is wrong or is seen on flat surfaces. wrong is untenable. It suggests a reluctance to The third expectation incorrectly assumes make peace with visual reality by allowing it that because a concave cube is not a convex to be what it is. It is not conceptual reality, the cube no picture can depict both at once. The reality that we imagine. expectation is dead wrong in this case, though With regard to the first expectation, is the error is rarely noticed. An outline draw- three-dimensional space a rigid grid in which ing of a cube that is convex is absolutely near and far have fixed positions? The answer indistinguishable from an outline drawing of is immaterial because the nature of three- a cube that is concave. One can be traced from dimensional space is not at issue in the cube the other, and the drawing, or diagram, can be illusion. Pictorial space, a flat plane, has two read either way. dimensions. Because anything flat, such as the The reading depends on which edge of the surface of a piece of paper, tacks depth, near cube is assumed to be farther away. If I want and far (which refer to depth) are arbitrary or neither end farther, I can read the picture as meaningless concepts. not a cube at all, just a hexagon with three lines To prove that near and far exist only as a radiating from the center to three angles or ver- matter of interpretation on flat surfaces, draw tices. Adding to the picture does not a straight line on a piece of paper. Which end eliminate the ambiguity. A rigid three- of the line is farther off in the distance? As no dimensional surface cannot be both concave distance exists on the flat surface, choose and complex from the same perspective. But whichever answer you want: (A) neither end, a drawing or projection of a surface of this type or (B) whichever you label "more distant." follows another set of rules. A single picture, The second expectation incorrectly like a single word,can have multiple meanings, assumes a one-to-one relationship between not just many shades of meaning. 150 Delusion and the Geometry of Visual Space

Impossible Figures world-objects we will never bump into. We In Surrealist art, optical illusions, sometimes all know that from seeing pictures of unicorns, combined with puns, are gateways to the angels, and dragons. unconscious for anarchists. They occur in a Escher's pictures of impossible figures, like formalistic manner in Op Art, displayed for our other artists' pictures of unicorns, are real in delectation as if mysteries to plumb. In a class that they fit the definition of real pictures. Each of illusion made familiar through prints by M. is a two-dimensional configuration. Their only C. Escher, pictures are drawn of objects that limit is that the marks on the planar surface could not be constructed in a three- depict what could not or does not exist in the dimensional world. They confute the laws of three-dimensional world. The impossible structural mechanics just as other objects that figures defeat the arbitrary expectation that can be drawn-unicorns, sphinxes, dragons, what cannot exist in three dimensions (and angels, and mermaids -- confute the laws of therefore cannot be integrated with our con- biology (figure 17-3). ception of reality) cannot exist in any form Irreconcilable with our tactile conception anywhere. of reality, the structurally impossible figures If illusion (or delusion) is a label applied to are linked at a remove with incommensurable that which contradicts expectations, the numbers, discovered by the Pythagoreans immediate pointing is to the provisional nature when it was found that lines could be con- of human knowledge. What we know never structed that could not be measured. The consists of more than what we think we know Pythagoreans were upset to discover that their at the moment. Picking and choosing our facts, expectations had been wrong. To their credit, we construct our private and collective pic- they did not argue that the unmeasurable lines tures of reality. Over the centuries, facts that (say, the diagonal of a square with one-inch did not fit, including many about color and sides) were illusions or not really there. vision, were disposed of by labeling them illu- Because the term is used arbitrarily, sions. This banished them to a conceptual Escher's impossible figures are called optical limbo of facts thought not worth considering. illusions and drawings of unicorns are not. The For Descartes (1596-1650), that objects drawings teach us not to expect a two- known to be large look small when seen from a dimensional world to match a three- distance was a delusion of the senses. Yet the dimensional world. They also teach that a two- mechanics of perspective had been under- dimensional universe is not subject to more stood in the visual arts for a hundred and fifty stringent limits just because it has fewer dimen- years before Descartes was born. Because sions. Pictures can be drawn of objects that perspective can be given an optical, not simply cannot or do not exist in the three - dimensional a mechanical, explanation (the explanation concerns how rays of light pass through a lens), the question is where the delusion lies. We cannot reasonably say rays of light are deluded when they persist in passing through lenses (whether glass lenses or the living lens of the eye) in one manner rather than another. The assumption that anything ought to look Figure 17-3. Impossible figure illusion. Just as lines that are incommensurable (impossible to measure) like "what it really is" is baseless, especially can be constructed, figures that are impossible to since our ideas are often unclear about what construct can also be drawn. things are. Equally baseless is that any object Delusion and the Geometry of Visual Space 151 ought to look the same under all conditions, three-dimensional reality. Visual space is a two- without regard to such variables as viewing dimensional world with a reality of its own. distance and changes in illumination, without Less sensational indicators of the unique regard, in short, for the relativistic aspects of nature of visual space are more profound than, visual space, time, and color. Not finding what say, the Hering illusion. Euclid's fifth axiom we expect, we feel comfortable concluding that implies that any given line has only one parallel something is wrong with our eyes. That passing through a particular point. There- something is wrong, instead, with our think- fore, parallel lines never meet. But parallel lines ing is more difficult to consider. meet, as anyone can see, in railroad tracks In the ideal world envisioned by receding to a horizon. In perspective drawings, Descartes-indeed, demanded as if it were the converging of parallel lines-that they can be superior to our deluded domain-nobody would nonparallel-is tacitly acknowledged. know whether it was day or night, because the Today a ray of light is often regarded as the colors of objects would remain unchangingly ideal straight line. Parallel rays of light are not "what they really are." Nobody would be able always parallel either. Illustrations in books on to judge distances, because all objects would optics diagram the diverging and converging unvaryingly look "the size they really are." And of parallel rays of light passing through lenses. we would all fry to a crisp. The sun, showing Space in Euclidean geometry (a space in which itself in this crowded world as large as it really parallel lines never meet) is not consistent with is, would for consistency be as hot as it really visual space. The inconsistency is not resolved is, or as hot as it would feel if not millions of in projective geometry, which undertakes to miles away. explain what occurs when shapes on a spher- I see few advantages in this transcendental ical surface are projected onto a plane. Part of nonsense world. Perhaps people could derive the ambiguity in the Euclidean system, which moralistic satisfaction from its putative prevents its reconciliation with the geometry revelation of "what things really are." They of vision, is the fixed orientation of its picture might be relieved to discover, at last, a world plane. Like the far walls of rooms in Vermeer's in which what they saw outside their heads paintings, the Euclidean plane lies perpendic- nicely matched up with ideas inside their ular to the observer's sight line and is heads. They could interpret this as a freeing presented as incapable of rotation in most from the delusion of the senses, a flight cases. upward to realms of higher knowledge. Yet the Geometry gives us our conception of three- only higher knowledge is that we cannot dimensional space, in which we believe, say, demand from the world a truth other than that that parallel lines are always parallel. They stay it presents us. the same (parallel under all conditions) and never meet. Descartes, regarding perspective effects as delusions, was assuming or asserting The Geometry of Visual Space the Euclid's fifth axiom is right and what we We cannot regard color thoughtfully without see is wrong. Why is space in geometry realizing that visual space, the space of fiat sur- inconsistent with space as we see it? Is Euclid's faces, is unique. This space has only length and fifth axiom wrong? Does it need further refine- width, and no depth. As optical illusions teach, ment? visual space cannot be understood in terms of After generations of mathematicians had rules that apply to three-dimensional space and puzzled over whether Euclid's fifth axiom was often fails to conform to our conception of more properly an axiom or a theorem (deriv- 152 Delusion and the Geometry of Visual Space

able from the four axioms that precede it) and whether it could be proved, it was discarded. The gesture opened the door to the non- Euclidean geometries. Carl Friedrich Gauss (1777-1855), Janos Bolyai (1802-60), and Nikolay I. Lobachevsky (1792-1856) devel- oped hyperbolic geometry. Bernhard Riemann (1826-66) developed spherical non-Euclidean geometry. Felix Klein (1849-1925) proved that Figure 17-4. A Euclidean point creating these and other alternate geometries were each a line. A line segment is the path traced logically as valid as Euclid's. by a moving point, which one must assume had an origin. Whether the The non-Euclidean geometries were not origin is in space-time or in "infinity" is inspired by the meeting of parallel lines at the irrelevant. vanishing point in perspective systems. Nor was the inspiration the meeting of, say, railroad plane is rectilinear. Irrespective of whether it tracks in the far distance in nature. The rail- has edges, it possesses vertical and horizontal road tracks and other perspective effects are axes at right angles to one another. They cor- interesting to artists and architects. Art had respond to the respective directions of move- traveled the same path as mathematics with- ment of the original point and the line derived out notice, relying on visual rather than mathe- from it (figure 17-5). matical reasoning. Attempts to represent This familiar means of developing plane perspective phenomena in art date back to the from line is unnecessarily complicated. Both Romans, long before development of the alter- ends of the line are assumed to move, yet only nate geometries. By Descartes's day, books one need move. The other can stay fixed. The were in print on perspective, and any compe- line's lateral movement is also logically incon- tent artist understood the system. The remain- sistent. We had intended to follow the travels ing question is whether Euclid's basic of one Euclidean point through space. The assumptions, not just the fifth axiom, can be line's lateral movement compels us to assume adjusted to more easily reconcile them to what we see.

Point and Line to Plane in a usual manner of imagining the foundations of geometry, a single point is thought to cre- ate a line by moving in space. Nothing need be assumed other than the point, its move- ment, and the direction of this movement (fig- ure 17-4). The difficult question is how that line develops into a plane. It might move laterally, Figure 17-5. A Euclidean line creating a creating a plane indistinguishable from an plane. An infinite number of parallel infinite number of parallel lines by lying along- lines lie alongside one another to form the Euclidean plane, as if the original side one another. Like the sheets of paper on line had moved laterally, with neither which we write our thoughts, this Euclidean end fixed. Delusion and the Geometry of Visual Space 153 an infinity of other Euclidean points, required to explain the infinity of parallel lines that form the plane. Each of these lines is distinct from the original line, in that it passes through no points on the original line. In one way of imagining the construction, the infinity of additional points is required to provide points of origin for the infinity of par- allel lines.In another way, the infinity of points are those through which, respectively, each member of the infinity of parallel lines passes. In either case, the infinity of additional points is inconsistent in a system that proposed to begin with one point and never explained why Figure 17-6. A non-Euclidean plane. If more were needed. More troubling, each of the line is assumed to remain fixed at its point of origin, its movement is limited the infinity of parallel lines extends infinitely. to circumambulating that point. The How can the lines all extend to a single plane created (or traversed) has infinity if they never meet? We are compelled curvilinear rather than rectilinear coor- to accept an infinity of infinities, a messy con- dinates. It can be more easily reconciled ception. with visual (phenomenological) events Imagine a different way of creating a plane. and with the rules of mechanical Only one end of the line moves. The end cor- perspective. responding to the point of origin remains per- corresponds to the respective directions of manently fixed. Movement, given these limits, movement of point and line (linear, radial). can only be radially around the point of origin, If the process of rotating the line around its as if the line were a spoke on a spinning fixed point of origin is extended to three wagon wheel moving radially around the hub dimensions, the volume created is spherical, an (figure 17-6). In everyday terms, an automo- outgrowth of both the fixed point that forms its bile (the moving point) starts out from center and the great circle of its cross section. Washington (the fixed point). The car drives Returning to the moving automobile, assume down a perfectly straight road as far as, say, that Washington (the fixed point) is floating in New York (any distance will do), and then outer space. The automobile (the moving point) decides to change direction. The car moves in has driven as far as New York, also floating in a huge circle that finally brings it back to New space. The automobile is tied to Washington by York. Washington is the center of the circle. a taut, unbreakable chain. The automobile, The distance from Washington to New York which can fly through the air, is given the task is the radius of the circle. of driving through every point in space that A plane produced in this manner is cur- will keep the chain taut. The volume described vilinear, again without regard to whether it will be a sphere. possesses edges.One of its "axes" corresponds The archetypal Euclidean volume is cubic to all possible radii of a circle. The other con- rather than spherical. The cube is created sists of an infinite number of concentric cir- when the Euclidean plane, that infinity of par- cles moving outward from the original point. allel lines, moves at a right angle to itself Whether or not axes of a coordinate system through an additional infinity of parallel lines in a usual sense, each of these respective axes (figure 17-7). Thus, a troubling aspect of geom- 154 Delusion and the Geometry of Visual Space

of the universe as we know it out of this singular point. I doubt anyone envisions the Urpoint moving due south to form a line; thence the line traveling east-west to form a plane; thence up-down to constitute a cubic volume, with the requisite x, y, z coordinates. If only because more consistent with what we know about explosions, the easier image is of radia- tion in all directions from a center.

Visual Space as a Circularity Parallel lines in railroad tracks meet in the dis- tance less because the senses are deluded than because the logic of Euclidean geometry, which directs many of our expectations about Figure 17-7. Euclidean volume (a cube). the nature of the world, falls short of provid- The prototypical Euclidean volume is ing a reliable picture. Irrespective of whether the cube, formed when the plane this is the sense in which space is said to be repeats the line's motion in moving laterally through space. Its infinite par- curved in theoretical physics, space as we see allel lines never meet, or they meet only it behaves as if it were spherical: a radiation at an infinity that is metaphysical from a fixed point, or vanishing point, toward because it can never be reached.

etry is its inconsistency on the question of how a cube (cubic space or cubic volume) can be created at all. If a Euclidean plane is synony- mous with an infinity of parallel lines, we can- not reasonably have more than this infinity of lines. Yet this excess of infinities must be assumed to extend the plane into a third dimension. The problem is avoided in creating a sphere because one end of the moving line remains fixed at its initial point of origin. Constructing a sphere requires an infinity of lines radiating Figure 17-8. Non-Euclidean volume (a from a fixed center. But we need neither more sphere). The sphere, the prototypical lines than this infinity nor fewer lines (figure volume for non-Euclidean space, is formed when the line, with one end 17-8). fixed, moves through all possible points While lip service is paid to Euclidean geom- in space other than this fixed point of etry, we ignore its tenets in imagining models origin. In spherical (visual) space, parallel lines meet at an infinity of the world. The big bang theory, for exam- identifiable as the center of the sphere, ple, proposes that the universe was originally the vanishing point of a perspective a singularity: a point that preceded time and system, or the point from which the space. We are asked to imagine the explosion Ur-line originated. Delusion and the Geometry of Visual Space 155 which receding parallels converge. That no allows parallel lines to meet at a vanishing tactile equivalent can be found for this visual point, might be regarded as an outward converging is immaterial. The reason I cannot projection of the spherical shape of the human obtain verification that parallel lines really eyeball and the focal point of its lens. Possibly, meet at vanishing points on a horizon is that in a world where intelligent beings had eye- my arms are not long enough to reach the balls shaped like cubes, Euclidean space would horizon. And the horizon recedes as I move be more consistent with visual space and closer to it. parallel lines would never meet. That world Vanishing points, which lie on the horizon would require different optical laws than ours, or above it, are assumed in systems of and I will not conjecture about what they perspective drawing. What is a vanishing point might be. (a point at which railroad tracks meet) in terms of the phenomenological world? Although more easily understood as an optical Nonconverging Parallels manifestation than through geometrical We cannot leave the subject of parallel lines analogy, the vanishing point is not an entirely without acknowledging the many occasions on subjective phenomenon. Just as I cannot see red which they fail to converge. Railroad tracks as blue, I cannot by effort of will relocate the viewed from an airplane flying above them look vanishing point from where it happens to be. parallel, which raises the question of how to But neither can I turn my back on it. Railroad explain the vanishing of the vanishing point. tracks that recede to a vanishing point in the Imagining it relocated above my head as I look east, when I look in that direction, will recede down from the airplane is not a satisfactory to a vanishing point in the west when I look the solution. It says nothing more than that the opposite way. vanishing point cannot be found within my field As if space were the ultimate fluid, or as if of vision. no privileged coordinate system existed, the In a better model, imagine an opaque plane vanishing point is always central to my visual lying perpendicular to the line of sight-the field, controlled by the direction in which I orientation assumed for the picture plane of look. It can even be more than one point, as in Euclidean geometry. This opaque plane blocks two-point and three-point perspective systems. the vanishing point from view by preventing our In the great sphere of visual space, any point, seeing the distance into which parallel lines anywhere, can function as a center depending might recede. Like Newtonian physics, on which way I look. Euclidean geometry represents a limited case, I conclude that the vanishing point, though the single case in which the spherical geometry prospectively any and all points in space, is a of vision is occluded by the Euclidean wall. single point. The likely candidate (because I If we always looked down on railroad can move it around at will) is the focal point tracks from airplanes, the tracks would never of the lens in my eye, projected outward as if meet at a vanishing point on the horizon. it existed in a far distance rather than on the Euclidean space would correspond to visual visual plane. The focal point of any lens is an space, and the fifth axiom would hold. Euclid- attribute of that lens. We need to look through ean geometry is excessively idealized, which the lens to be aware of it. Yet the focal point is why its concepts confused Descartes. I find never appears to reside in the interior of the it odd that this also means Euclidean geome- lens. try gives its best explanation of the world we The spherical nature of visual space, which see when applied looking down from air- 156 Delusion and the Geometry of Visual Space

planes. Its conception of space no longer works Might we, reasoning from this, develop a when the airplane lands. new geometry that began with the colors we Everything we see, as has often been said, see; that assumed straight lines are paths presents itself as both color and form. Yet traversed by rays of light, and that moved on to conceptually, or through long habit, we separate deduce the rules of geometry as we know them? one from the other, relying on geometry to tell This would be a strange kind of geometry, us about shape and perspective (regarded as blending phenomena we now regard as objective) but not about colors (regarded as psychological with others we now regard as subjective). The distinction is artificial. The mathematical. Or it might be regarded as a horizon and the vanishing points of perspective joining of two kinds of geometry. The Munsell systems (and of perspective geometry) have no and Ostwald systems for -organizing colors, phenomenological reality as objects that can be like many simpler systems that preceded them, touched. Both, like color, are subjective, envision color solids that are geometrical con- two-dimensional manifestations never exactly structions. If color relationships can be the same for two observers or two cameras. The explained, at least to an extent, in geometrical horizon is a color phenomenon, perceivable terms, then why cannot this explanation be only as an abutting of the colors of the sky with incorporated into Euclidean geometry or into those of the earth. any of the several non-Euclidean geometries? CHAPTER 18 Extension in the Visual Field

As you ramble on through life, brother, Whatever be your goal, Keep your eye upon the doughnut And not upon the hole. Mayflower Coffee Shops

I olor is unique to the visual field, in color, as does the doughnut itself. The colors that it cannot be seen perceived any- of the hole are attributable to the plate, table, where else. it is also the basic element or other objects lying behind or beyond the in the visual field, Images of objects are made doughnut. Although conceived of as empty in of color. So are images of the spaces in and three-dimensional terms, holes never look around them, a difference between what we empty in the visual field. Color uniformly touch and what we see. In three-dimensional permeates the images of both objects and the terms, a doughnut from a bakery shop is a spaces between them. Everything we see, as three-dimensional object. Its hole is a volumet- has been repeated, is some color or collection ric space that is empty. The two-dimensional of colors. world of visual imagery has no doughnuts, but That the visual field is devoid of holes or it has images of doughnuts. We call these empty spaces means we cannot see them as images doughnuts as an expediency, and the such. A visual gap, by its nature, is visually expediency sometimes causes confusion. imperceptible. We need not, in pondering this mystery, rely only on the example of the nonemptiness of images of doughnut holes. A Holes and Spaces blind spot is known to exist at the center of Visually, the hole in the doughnut (or in any the retina. The evidence, which is compelling, image of a doughnut ) looks like an area of is that no rods or cones are found at the junc- 158 Extension in the Visual Field

tion of optic nerve and retina. But the eye's image, the limitation that defines imagery as blind spot is invisible: without phenomenology- two-dimensional. cal consequence as a visual field phenomenon. I sometimes think the idea that breaking a What we see has no gap, hiatus, or empty place mirror brings bad luck is promoted to in its center. The visual field, in other words, discourage inquisitive children from smashing does not ordinarily give the impression of mirrors to determine whether this is the way to looking like a doughnut. get inside. Expectation is offended by the A geometric or topological logic underlies discovery that images have a front but no back, the absence of holes in the visual field, causing an outside without an inside. As human beings, the all-pervasiveness of its color. Holes always we remain forever outside the image, in front of go somewhere, whether into things, under it and never behind it. Or images are always in them, or behind them. In the three-dimensional front of us. We lose touch with them if they world, holes are experienced as such because move behind us or to any other location. I things can be put into them, an experiential cannot see what I am not looking at. I cannot confirmation of their emptiness. Doorways into look at what is not in front of my face. a third dimension, holes remain for this reason The obdurateness of surface is not confined inconceivable (and also imperceptible) in a to the surface we know as the visual field. planar, or two-dimensional, realm. Nothing lies What we are able to touch, in the three- behind the images of the mirror, the eye, the dimensional world, is confined in most cases to photograph, the movie film, or the picture the curved planar surfaces of objects. These plane of a painting. A hole in the surfaces, because two-dimensional, have little two-dimensional continuum of any image has in common with the volumetric forms that no place to go. they cover, which in a sense cannot be The three-dimensional world in which reached. Cutting an apple in half is less a jour- volumetric forms are located is a discontinuous ney inside the apple than a creation of addi- domain of objects and empty spaces. The tional exterior surfaces, a revelation of further two-dimensional realm of images is both aspects of the apple's image. Like the surface immaterial and homogeneous in this of the outside of the apple, the surface of the immateriality. As Maurice Denis remarked inside is colored, in this case not in the same about the picture plane of a painting, images in colors. general are aggregates of color spots. If not recognized as such, they are misunderstood as windows in a wall. The viewer assumes an The Extension of Color and Form imperative to look into, or through, the Within the visual field, the most striking imagined window toward an imagined third characteristic that color and two-dimensional dimension, a charming but insubstantial idea. (visually perceived) shapes have in common The colored images of objects cannot be is dimension or extension, the ability to vary pierced to gain access to any metaphysical in age and size. Objects, including perceived essence hidden behind the skinlike surface. color spots or aggregates of color spots, Whether the images are those of the visual field become older through extension in time. They or of the picture planes of mirrors, photo- become larger through extension in space. graphs, and paintings, we cannot look through Spatial extension is reversible in theory. Any- a two-dimensional configuration because noth- thing that expands can contract. Temporal ing lies behind it. Anything behind an image, extension is more complex. In theoretical like anything in front of it, is not part of the physics, Feynman diagrams show electrons Extension in the Visual Field 159 and other particles moving backward and related idea of a world endangered by forward in time. But in the everyday world, something devoid of constraints on its what has been called the arrow of time points in extension. The malevolent (or Malthusian) one direction only, a message conveyed by the animal, plant, insect, virus, rock, or substance nature of death. All life moves onward to die, threatens to grow larger or reproduce itself (a and the dead do not return. The future, in way of increasing) until it crowds out forms of simpler terms than ordinarily used to define it, life other than itself. The Incredible Shrinking is when we will die. The past is where those Man, a classic science-fiction tale made into a already dead can be found. movie, imagines extension in a reverse The changes we explain by attributing them direction: the protagonist shrinks smaller and to the passing of time may need a better smaller, apparently ad infinitum. explanation. To ask whether time moves forward or backward in the everyday world is meaningless. No criterion exists for the Extension as Becoming difference. In vision, as in the corporeal realm, Extension is dependent on becoming, which change or the passing of time cannot be halted. implies absolute fluidity. Its implied opposite Among visual pathologies, both physiological is being, which suggests absolute stasis. and psychosomatic, no such thing occurs as a Although the objects of the everyday world person afflicted by a selective inability to see become older, larger, or smaller, Zeno's para- motion. Cinematographers can freeze a single dox can be expanded to argue that extension, frame and theoretically hold it forever. The like motion, is impossible. No movement functioning eye cannot similarly escape time, exists if we merely rest at various points in which behaves in vision as in the world of time. Nor, by a similar reasoning, can objects three dimensions. Either we see in terms of become larger or smaller. They are just one time (the forms that are seen move), or we do size at one point in time, another size at not see at all. another. Every object appears to have constraints to Pursuing this reasoning leads to a model of its potentiality for spatiotemporal extension. the universe based on a reel of motion picture Human beings rarely grow taller than seven film. Each frame, separate from any previous, feet or live longer than a century. Jain is an entirely new picture. The world would be mythology attributes a spiritual significance to in a unique state of being every instant, never these limits. In the first descending period of the same as a moment ago nor connected to any the Jain cycle, human beings are said to have worlds of the past. This vision is the ultimate been six miles tall; they had 256 (or 28) ribs. existential extension of Hooke's argument that In the second period, their height was reduced Newton's prism manufactured the spectral to four miles, their ribs to 128 (27); in the third colors. period, two miles and 64 (26) ribs. Progressive Despite its intuitive appeal and precedents shrinkage reduced human beings to the size of in religious and philosophical thought, this our era. In future, people will shrink to vision has rarely commended itself in the West. eighteen inches, have 8 ribs, and live fewer Stability cannot be presumed for scientific laws than twenty years. After this, the order of the unless inherent in the world to which these epochs will reverse. Men and women will laws apply. From a scientific point of view, we grow progressively larger, with more ribs and find it expedient to think of the world as longer life spans. always the same (or always the same even A familiar science-Fiction plot toys with the though it changes) , rather than continually 160 Extension in the Visual Field

different. When regarded as immutable despite on the camera. Pictures made with fish-eye change, the world acquires a history, a record of lenses often include the photographer's feet but the past and a key to predicting the future, a never the photographer's face. talisman against terrors and uncertainties. The Mirrors, which the Etruscans and other cost of the talisman is loss of sensitivity to those ancient peoples buried with the dead to ward aspects of the experiential world difficult to off evil spirits, partially transcend these opti- conform to the system. cal limits. I can look at myself looking at myself Assuming that extension exists, the exten- in a mirror. Mirrors are preferred as light- sion of the visual field is limited. Unless I turn gathering devices for large telescopes precisely my head, change location, or use a mirror, I because they do that work more efficiently cannot see objects behind me. Nor can I see than lenses. Even a mirror is unable to see what behind objects in front of me. Constrained by lies behind itself, an insurmountable physical the human condition, I see the objects in front limit. A sphere with a mirror surface, reflect- of my face and only that portion of the sur- ing everything around it, is unable to reflect its face of each object that happens to be facing own interior. me. Axiomatically, nobody can see what is not The limits of the visual field are measur- being looked at, and looking at everything is able and fairly stable for any individual. impossible. Something always exists that is not Measuring from the pupil of the eye, the being looked at, a limit integral to lenses and average human being sees about sixty-five mirrors, not a peculiarity of the human eye. degrees on the nasal side, ninety degrees on the Traditionally, the ability to see everything, or temporal side. In a small number of subjects, the everything at the same time, has been size of the visual field varies from day to day, attributed to God or the gods, probably for some reason that has not been determined because no ail-seeing eye, lens, camera, or mir- (Fuchs [ 1908] 1924, 256). In hernianopsia, ror exists. Why do the photographers' faces caused by pituitary abnormalities, half the field not appear in pictures taken with 360-degree for each eye ceases to function (Fuchs [1908] fish-eye lenses? Why do we have to stand in 1924, 263). front of a camera, not behind it, to take pic- The fish-eye lens and other wide-angle cam- tures of ourselves? I suspect that the limit could era lenses have a more inclusive field of view be expressed in mathematical terms and could than the human eye. But, just as writing an be shown to relate to the number of dimen- exhaustive autobiography is impossible sions in our world. Transcending it might because the author will have more experiences require a world with a greater number of after the autobiography is completed, for any dimensions, in which, say, a circle could have eye or lens to see everything in its surround- more than 360 degrees. From our world, we ings is impossible. Optical and conceptual know only that the 360 degrees swept by the difficulties prevent our even imagining an eye fish-eye lens does not reach everything. or lens with such an all-encompassing field of Images have other limits that refer directly view that it could observe itself observing to dimension. Images in a mirror, whether itself. regarded as illusions or as reflections, refer to The 360-degree fish-eye lens is as con- a three-dimensional world but are planar (two- strained as any other. We cannot see what we dimensional) themselves. Volumetric space is are not looking at, and the fish-eye lens is not unable to function as a mirror of an analogous looking at itself. Nor is the lens of a human eye kind, for no three-dimensional mirrors exist that may be looking through the fish-eye lens that are capable of reflecting four-dimensional Extension in the Visual Field 161 space. Nor do we have one-dimensional mirrors The misestimation has a subjective aspect. As that reflect two-dimensional space. the Belgian Surrealist Rene Magritte learned from Hieronymus Bosch, a sense of the uncanny is created when objects look greatly Extension and Size out of scale with the size we expect them to In tactile terms, the perceived size of an object be. is the size felt by a moving hand. In visual Like the Ames room, the depth perception terms, because of the phenomenon of perspec- experiments of James J. Gibson shed less light tive, a translation is needed. The extension, or on the nature of perception than on rules for perceived size, of any object seen by the reasoning about what is perceived (Gibson unaided eye is accounted for by two factors: the 1950, 1968). If the distance of an object is real size of the object and its distance from the incorrectly estimated, an inversely related viewer. From ten feet away, one person may mistake about its real size cannot be avoided. look larger than another because he or she is The ancients Greeks believed the fixed stars larger. But a person five hundred yards in the were small and relatively close to the earth. We distance looks smaller than another six feet believe them to be large and at enormous away, regardless of the size of the people. distances. At issue are differing game plans for Our estimates of the physical size of objects rationalizing extension: for balancing the books begin with extension, the perceived size of the by explaining the size that the stars appear to be object relative to other objects in the visual in the visual fields of human beings looking at field. An assessment is made of the degree to them. which visual size is determined by distance, We have no subjective sense of the absolute and of the degree to which visual size is deter- size of a visual image on the retina of the eye, mined by the actual dimensions of the object. which must be exceedingly small. If I saw the For very distant objects, color and brightness page of a book no larger than the fraction of an provide clues to distance, though the clues are inch the image occupies on the retina of my eye, neither easy to read nor reliable. A distant the print would be too small to read. Because a mountain appears veiled by blue haze to a sense of the absolute size of visual images on greater extent than another mountain nearer the the retina is absent, a sophisticated algorithm for viewer, but the amount of haze varies from day juggling ratios is likely involved in estimating to day. Distant stars look more dim than those sizes and distances. The process never intrudes nearer, but a bright star can be farther away on consciousness as mathematical, and animals and larger than a small, dim, nearby star. judge sizes and distances that interest them as Because the environment provides visual competently as human beings. clues to size, input into the system can be The microscope and the telescope increase manipulated. The clues can be tampered with. the visual size of objects, their extension The colonnade with false perspective by the within the visual field. The microscope Baroque architect Francesco Borromini creates enlarges objects otherwise too small to be an illusion of greater depth than is present. seen. The telescope enlarges those too distant The Distorted Room, constructed by the (therefore, too small) to be seen. In either case, American psychologist A. Ames, tempts the lenses and curved mirrors supplement, or observer to wrong conclusions about the real increase, the focal length of the lens in the eye sizes of objects (Gregory 1966, 178-179; Ittel- of the human being using the instrument. The son 1952). Its playing with perspective invites mechanism suggests that the experimenter incorrect estimation of viewer-object distance. cannot be separated from an experiment. 162 Extension in the Visual Field

Whether anything is seen through a microscope mathematical bases and constraints. We do not or telescope, and what is seen, depends on the know how to read the signs, how to arrive at a lens of the eye looking through the instrument, sufficiently sophisticated understanding of a vital part of the machinery when an what we see. observation is made. The limit is not overcome The conception of an arrow of time, in the by replacing the eye with a camera because a physical sciences, expresses the idea that time human eye is still required to look at what the moves in one direction only. An arrow of camera records. vision exists and is also one-directional, though Enlarging is not always sufficient, or this second arrow receives no attention. I can required, when an object is difficult to see. To look in only one direction at one time. I cannot see large bodies in their entirety, a mechanism be inside a system looking out at the same time is needed to make them look small enough to that I am outside the system looking in. I fit within the visual field within a single cannot be behind a camera lens and in front of moment. Astronomers and cosmologists, con- it at the same time. Photographers' faces do not fined by this limit, can offer no more than con- appear in photos taken by fish-eye lenses for jecture, however sophisticated, about the shape the same reason that the earth when seen from of the universe. Direct visual verification, if its surface looks different than when seen from possible in any ordinary sense, could be outer space. The issue is the arrow of vision, provided only by an individual observing the which allows us to look into a coordinate universe from its outside, from a sufficient system in which we see as far as the vanishing distance to see the entire system at once. point of that system. I can choose to look east. Because systems cannot be seen from the I can choose to look west. I am not given the outside by an observer located inside, human choice of looking east and west at the same beings had no way to confirm the roundness of time. the earth visually until the twentieth century. Astronauts see this roundness from outer space, and we see the photographs they took. The Edge of the Visual Field Descartes's opinion to the contrary notwith- Color and form display dissimilar attributes standing, delusion of the senses was always a when extended to the edges of the visual field. poor explanation for why the round earth looks Color behaves as if it were a substance rather flat to creatures living on its surface. The than an object. In an important visual dynamic, barrier is topological or perspectival and relates seeing nothing but a single color is possible, as to the nature of coordinate systems. in seeing the darkness of the night. We cannot How any object, including the earth, looks see, or imagine seeing, nothing except a single depends on the perspective from which this form. object is seen. We cannot reasonably expect The behavior of two-dimensional forms in that an object viewed from a point on its sur- planar universes differs from that of three- face will look the same as it would if observed dimensional forms in volumetric space. In fig- from a point beyond or outside that surface. ure 18-1, a small black disc grows progres- Different conditions-different points in a sively larger in its white field. At maximum coordinate system-suggest differences in how extension the disc,which fills the field entirely, things look. Rather than deceiving us, vision is no longer identifiable as such; it is now a probably provides very keen insight into the black rectangle. No visual clues exist to its nature of the phenomenological world, its original nature, which can only be understood Extension in the Visual Field 163

the separation of light and darkness to tell about the creation of a multitude of forms and spirits. We are not told that God created colors, numbers, or himself. Nor are clues provided about how any member of this three- some came into being. Either all three always existed, or another creation story explains them. Why are we not told how these three items were created? I see a relationship among them on two levels, a relationship that might have interested mystics, although I know of no evidence that it did. On the first level, the darkness of the primordial void is unimaginable without a prior assumption of color (through which its extension might have been perceived) and Figure 18-1. A form growing larger in a field. A number (by which it might have been form that expands to breach the perimeter of its field measured). Although we never think of the merges with that field, or becomes indistinguishable writers of the Old Testament as logicians, a from it. The form cannot be measured "within" the field. good logical argument might be made that color and number are first things, the necessary prior assumptions that allow us to assume the by knowing its history. In effect, it no longer primordial void. How would God have known has that nature, which was simply a temporary the void was there had he not been able to see it attribute. (color) and take its measure (number)? Although the black rectangle gives no At a second level, relationships among the reason for suspecting it might have been part of three uncreated elements point to the kind of a circle, a person asked to draw a rectangular paradox that delighted medieval theologians. segment that is part of a circle would be Or they suggest the children's game that asks compelled to arrive at the figure of a rectangle. whether rock, fire, or water is strongest. Is A form extended to fill any planar field, or to number greater than color? Is color greater than exceed its limits, is no longer identifiable as the number? Number, we might say, exceeds original form. For this reason, the black circle color, because we cannot know the color of is no longer a circle in the final drawing of numbers. But color exceeds number, because figure 18-1. Its color, which has undergone no we cannot know the number of colors. The corresponding change, remains black deity of the Old Testament, limitless and throughout. In two-dimensional universes, superior to all else, fits nicely into the picture, form behaves in extension as a variable, or is perhaps too neatly. Because immeasurable, metamorphic. Color behaves as a constant, or is God exceeds number; because invisible, his immutable. "colorlessness" is similarly absolute. In Traditional intimations that the constancy another way of understanding the linkages, of color holds in both time and space can be Genesis presents us with two voids, the clas- accounted for by the difficulty of imagining sical antithesis between matter and spirit. The any other condition. Genesis proceeds from void containing darkness and light, from 164 Extension in the Visual Field

which the world was made, is matched by the leads to several visual consequences. Among limitless void which is God himself. them, the overall character of a form (unlike Genesis probably hides clues we no longer the overall character of a color) cannot be understand about the reasoning that led human determined from small parts of the original beings to conclude that a colorless, limitless unit. The fable of the blind men and the ele- God existed. If this God was discovered by phant strains too hard at a point that could have analogical reasoning, the analogy was with been made economically. One blind man feels empty space. Whether in Genesis or in the the ear of the elephant; another, its trunk; the modern physical sciences, human beings find third, its tail. Each generalizes from limited it easier to imagine empty space giving birth experience, arriving at wrong conclusions. The to forms than forms giving birth to empty three blind men are unable to agree about what space. Thought moves more easily in that an elephant looks like. direction for two reasons. First, space, which The protagonists of the story need not be has only extension, is a simpler element than blind. Sighted observers are prone to similar matter, which has both extension and mass. error if similarly unfamiliar with entire Second, empty space is needed to allow elephants. An elephant's ear, which does not motion, including the gesture of creating a feel like an elephant, also does not look like an world. No creation myth exists that declares the elephant. Details of a form, even visually universe was originally, say, a solid block of perceived details, cannot convey complete stone (or, for that matter, a spot of color). In information about the form as an entirety. The Buddhism, the idea of forms emerging from Egyptians and other early peoples showed a empty space is carried to the next logical pla- strong preference for depicting human beings teau: the forms, we are told, are illusory, no and animals in their entirety in art. The prefer- more substantial than the space from which ence had an informational aspect. Most they arose. animals are bilaterally symmetrical, allowing Color, like space, has extension without us to assume that the left side of, say, an mass and might be thought of as a visual cor- elephant's head will look like a mirror image of relative for the tactile experience of feeling the right side. No further axes exist, and nothing. The link is noticed by modern nobody ever learned what an elephant looks writers, more often by those writing on color like from seeing only its head. than by those writing on space. It rarely comes As if there were a geometry of vision, the to the surface in myth and legend. No creation limitation has a structural or topological story tells us, say, that the creator deity was a dynamic. Figure 18-2 shows what occurs when large blue spot. Genesis comes closest to the expanding black circle of figure 18-1 dis- dealing with color through its two voids: the appears into its own field. When the circle invisible, colorless void which is God and the passes beyond the perimeter, it can no longer other void we imagine as black, or black and be measured within that field. If we assume white. that the black field of figure 18-2(A) is part of a large black circle, no criterion exists for determining whether the circle is as shown in Fragmentation 18-2(B) or as in 18-2(C) or otherwise. A blue spot that grows larger loses its original Breaching of the perimeter is the significant size but can retain its original color. The factor, irrespective of whether the field is behavior of color as a constant in extension entirely filled. In figure18-3,a black object lies and of two-dimensional form as a variable partly inside and partly outside of a white field. Extension in the Visual Field 165

A B Figure 18-2. A form disappearing into its field. If limited to the confines of the field, we have no criterion for judging the nature of what lies outside. Noncongruent forms may appear visually congruent.

An observer inside the white field, unable to details, of, say, blue, green, or lavender. We see beyond its edges, cannot determine the think of forms, but not colors, in terms of entire surface area of the black object. The observer forms and details, topologically different has no way to discover the shape or size of categories. If a form is larger than the field in those parts of the black object that are not which the form is being viewed, what appears in accessible, that lie outside the white field. that field is a detail. Only part of the form is Figures 18-3(B) and 183(C) show two seen, and it may not look like the entire form. possibilities from a limitless number. We see details of the universe when looking Although this is rarely taught as a through telescopes, not the entire universe at mathematical rule, areas cannot be measured once. The cosmological principle, which is unless their edges can be located. Nor can actually a cosmological assumption, takes a volumes be computed unless their surfaces can bold conceptual leap. It tells us to assume that be found. This is why we have a more exact the universe looks pretty much the same in all knowledge of the volumetric size of the earth parts. How could anyone be sure of this unless than of the volumetric size of the universe. all the parts were seen? Nobody can predict how large a horse is, or what it looks like, from seeing its black tail. But if the horse is known to be a single color, we may correctly infer from the color of the tail that the horse is similarly black. If an area of color is a homogeneous blue, any fragment is as blue as the whole. Extension has no effect upon color. Nor does fragmentation, reversing the process of extension by dividing an area of homogene- ous color into smaller areas. Accommodating Figure 18-3. Form and area. We cannot measure an area outside a field from within the confines of that this constancy of color, the English language field because the edges of the area to be measured does not encourage referring to the parts, or cannot be located. 166 Extension in the Visual Field

Color in Fragmentation explained in this manner, assume that blue and The division of a drop of water cannot make it violet rays, known to have a greater power to less watery, a constancy that holds, in theory, penetrate deep into water, also penetrate far- until the molecular level. The division of a blue ther than other rays when moving through air. color spot similarly yields two blue color spots, We could then conclude that although objects not two spots each half as blue as the original. reflect rays of all colors, only the blue or vio- But no molecules of blue exist, to be reduced to let rays reach a distant observer, the reason dis- atoms of something else. And we are not tant objects look bluish gray. Carrying the encouraged to ask how long the process of thought further, the rays reaching a distant dividing a blue color spot can be continued. A observer may be predominantly ultraviolet, tiny particle of blue paper too small to be seen which has a shorter wavelength than violet, by the unaided eye would still look blue, I am looks bluish gray, and ought to have a greater sure, if examined by a microscope, as blue as it penetrating power than the spectral colors if would continue to look if blown up to fill the penetrating power is correlated with wave- universe entirely. length. Beyond the lower limits of the microscope, Despite these correlations, I am dubious the question is whether, like magnetism and about the wisdom that color cannot exist in thought, subatomic particles, presently known objects smaller than the wavelength of visible to us through their tracks in cloud chambers, light. The sticking point, as often, is the status are to be imagined as absolutely inaccessible of black and the blackness of the night sky. We to vision. Any possibility of seeing them assume that we are able to see nothing, the directly implies seeing their colors, because name applied to the black color of interstellar anything that has no color cannot be seen. If space. It would be surprising if the "some- the feat is a technological possibility, the eas- things" imagined floating in the nothing-the ier assumption is that the photons responsible subatomic particles suspended in the for blue might look blue, rather than some vacuum-were not similarly colored, even if other color. No basis exists for conjecture only similarly black in color. about the colors of electrons, protons, or If a portion of the blackness between the neutrons. stars were repeatedly subdivided, would it Few models exist to guide us on how to cease to look black? I see no way in which the reason intelligently in this direction. Because wavelength of light can have any bearing on color is correlated with wavelength of light in the question. Black is not correlated with the physical sciences, conventional wisdom in wavelength and is called an absence of the sciences holds that color cannot exist for wavelengths. Assume that a portion of the anything, say, an atom, that is smaller than the blackness of the space between the stars would wavelength of visible light. We need to continue to look black, no matter how often assume, in that case, that violet objects can be divided. Would the same be true for a naturally smaller than red objects, because violet has a black substance, say, coal? shorter wavelength. In favor of the assumption, We cannot, at these frontiers of conjecture, colors fade out in the depths of the ocean in this rely on such conceptions as the threshold of manner: violet rays penetrate deeper than those vision. The microscope offers evidence that of other colors, and the ability of rays of the although color is seen, it is more than some- spectral colors to penetrate ocean water is thing seen. Even exceedingly small objects correlated with wavelength. have colors of their own, revealed when the A similar phenomenon may occur in air. microscope brings them into our field of Although aerial perspective is not customarily vision. CHAPTER 19 Complementarity in the Visual Field

Objects are colored shapes, but one perceives the shapes only because they are colored. Charles Edward Gauss, The Aesthetic Theories of French Artists

s is familiarly noted, anything that can If form and color are complements, this be identified as a visual element can might be represented as in figure 19-1, mod- be reduced to an aspect of color or eled after the Venn diagrams used in symbolic form or both. Brightness is color variation. logic. In the elaborate conventions for reading Lines are the edges of forms or occur within diagrams of this type, the area of the square them. Motion is an event that happens to inside the circle (identified as form) is the forms, a meaningless concept in a formless complement of that outside (identified as color). void. The exploded view reveals the further Should color and form be regarded as com- relationship that is critical: within the visual plementary classes within the visual field? The field, nothing can be both a form and a color if concept of complementarity, derived from form and color are complements. logic, tells us that complementary classes are A substantial barrier exists to reasoning far- disjoint and exhaust their field. This implies ther in this direction. Every visually percepti- three essential conditions. First, no item can ble entity, with a single exception, manifests belong to both of two complementary classes. itself as both form and color at the same time. Second, every item belongs to one or the other Arnheim observed that a form is visible class. Third, the sum of the two complemen- because its color differs from the color of its tary classes equals the universal class. environment ( Arnheim 1956, 222). The 168 Complementarity in the Visual Field

Figure 19-1. Form and color as complements. If form and color are complements, every visual experience can be classified as one or the other, and no visual experience can be both. printed shapes of black letters on a page in a equate one with the other, the distinction has no book are seen because they differ in color from correlative in vision. When we speak of seeing the white page. A form indistinguishable in a spot of color, that spot is taken to have a color from its environment is visually discernible form or shape, the attribute that imperceptible as a form, which is why we allows it to be identified as a spot. Colors, cannot see objects in the dark. In figure 19-2, a however, are not invariantly perceived as spots. black circle, visible against a white background, The exceptional case is the color field. In this cannot be seen against a field of similar black. phenomenon, color appears dissociated from Isolation (or nonisolation) of a two-dimensional form, though visual form cannot appear without form from its surroundings is a function of its color. color. The color field is an area of any single That every visually perceptible form is dis- homogeneous color extending as far as the tinguished because of its color amounts to say- observing eye can see at that moment. This ing that any two-dimensional form "has" a translates, in experiential terms, to the color, "is" some color, or "is colored." Form experience of seeing nothing but, say, light and color are not mutually exclusive and there- green. The most familiar color field is the dark fore cannot be complements. Although no dic- of the night if the stars cannot be seen, or the tionary would define a form as a color or similar blackness of an enclosed, darkened

Complementarity in the Visual Field 169

mon in nature than seeing colored objects. But we can extrapolate by relying on either of two analogies. A red color field might be thought of as similar to the darkness seen at night, although it is red in color rather than black or gray. A red color field can also be thought of as the homogeneous red of a red color swatch, extended as far as the eye can see. Models of color fields are easily con- structed, and the term is familiar from critiques of color-field painting. A sheet of red paper Figure 19-2. Form and field colors. A black form, provides a convenient model, a miniature of visible against a white background, is invisible a red color field. The sole limit of the paper against a field of a similar black. is its scale. The color of the paper ends at the edge of the paper, not the edge of the visual room. No method exists for remaining in a field of a viewer. What occurs when we see, darkened environment yet also seeing around or imagine seeing, nothing except a single the edge of the darkness. In a parallel color? phenomenon, no way exists of seeing around, For a color-field experience to occur, or behind, the edge of the image in a mirror. stimulation must be spatially uniform and Color field conditions can be simulated in temporally continuous. If, during an interim, experiments. In nature they are approached in a what I see changes from all red to all yellow, group of rare visual pathologies collectively then I have not seen a red color field known as the chromatopsias. Similar in effect continuously in that interim. What has been rather than in cause, the chromatopsias lead to seen instead is a red color field for part of the discoloration of the jellylike humors inside the time, a yellow field for the rest. eyeball. The result is to overlay the visual field The number of possible color fields is with abnormal hues, although the condition is limited only by the number of possible colors. self-limiting by its nature. For vision to be The statement "I see nothing except . . ." can be possible, the humors in the eyeball must remain completed with the name of any color. The transparent, and any color they assume must statement is meaningful and unique irrespective remain transparent. of the color name used. The experience of In erythropsia, the field of vision looks red. seeing nothing but yellow differs from seeing A white wall would present that color, as if nothing but blue, brown, purple, silver, or being viewed through a pane of red glass. In white. Each color is capable of forming a xanthopsia, caused by jaundice, the overlay perceptible field of its own unique color, which color is yellow. In cyanopia, it is blue; in we may see, imagine, or model. ionthinopia, violet; and in chloropia, green. I can imagine seeing nothing but mauve, Several ophthalmological explanations are the color field that is exclusively mauve given for the chromatopsias (Fuchs [ 1908 ] extended to the point where it obliterates 1924, 247). anything else I might see. I cannot, to show the Apart from the darkness, which is a genu- limits of imagination, conceive of a color field ine color field, and the overlays of the chro- that, at one and the same time, is both entirely matopsias, which approach the condition, the mauve and entirely vermilion. Vision has con- experience of seeing colored fields is less com straints, many noticed by Aristotle, which we

170 Complementarity in the Visual Field

cannot surmount. These constraints are assumed verse, however, consist solely of transitions but rarely discussed, and they bear on theories from one color to another, as in a meeting about color. between redness and blueness. Although, say, every part of the visual field I conclude that two-dimensional form is an is colored, no part can be more than one color attribute of color, though rarely explained in at the same time, the reason a color field can this manner. We perceive spots of color and not be both entirely mauve and entirely infer that some spots can be understood as vermilion. Seeing color without form is a forms. Color is the percept; form, the possibility, as in the color field. Seeing form interpretation. Form (or forming) is an event without color is not possible. The tactile sense that happens to color, paralleling the manner in fills the gap by allowing us to touch objects in which motion (or moving) is an event that the dark, though we cannot identify their colors happens to forms. No motion can exist without by touch. If the laws of the physical world forms, and no forms can be seen without colors. allowed intelligent beings to exist who had no In visual terms, colors create forms and are not tactile sense, these beings would be unable to applied to them. The phenomenon can be see objects in the darkness and would be unable demonstrated by using models of color fields. to touch them either. COLOR FIELD MODEL I If two-dimensional (visually perceptible) forms Form as Discontinuity are disturbances of color fields, this implies Color is capable of assuming two modes. It can they ought to be regarded as events, as patterns occur as either a color field (in which the color of the moment. It also implies they can be is continuous) or colored forms (in which the created through work: the physical act of color is not continuous). Because nothing can disturbing. A sheet of uniformly red paper is an be both continuous and discontinuous, no adequate model of a red color field. It differs perceptual experience can consist of both at the from the largest possible field only in its scale: same time. If I see nothing but continuous in the observer's ability to see what lies beyond yellow, I cannot perceive, at the same time and its edges. place, a yellow disc situated in blue, gray, or Imagine that this paper is wrinkled into a multicolor surroundings. ball. It is then spread as flat as its wrinkled Forms, if this line of reasoning is continued, condition allows. Applying force to wrinkle can be defined as color discontinuities within the paper constitutes the performance of work. otherwise homogeneous color fields. They In its "worked over" condition, the paper no disrupt the continuum of uniform color (or of longer looks flat. Its surface is covered with formlessness) that would exist in their absence. wrinkle forms that look like minute hills and Seeing forms requires seeing more than one valleys. color. The color of the paper is affected by crea- The model of color in the visual field as tion of these forms. Color discontinuities have either continuous or discontinuous is similar to been introduced into the surface. If they were the classical conception of the three- not there, the forms of the wrinkles would not dimensional world as a discontinuous domain be seen. Although the paper is still properly of substance and emptiness: of corporeal forms described as red, the red is no longer uniform. interrupting what would otherwise be the con- Parts of the wrinkle forms are in shadow. Tak- tinuousness of the spaces surrounding them. ing readings from an ordinary photographic Discontinuities in the two-dimensional uni- light meter will confirm that the shadow areas

Complementarity in the Visual Field 171 are darker (a darker color). Other parts of the films rear-projected on screens. When the film wrinkle forms catch the light and are lighter (a is released, audiences rarely notice that they are lighter color). seeing photographs of the actors and actresses A more sophisticated test is to paint a pic- combined with photographs of photographs of ture of the wrinkled paper, recording what it the scenery.' looks like as faithfully as possible. The task Disturbing the black continuum of the movie cannot be performed with a single color, not screen to project a motion picture parallels even a color exactly matching the red of the wrinkling the sheet of red paper. In either case a original, unwrinkled paper. At minimum, a field of homogeneous color is disturbed by the second, darker color is required for those areas application of force: by work in a physical of the wrinkles in shadow.Or a second, lighter, sense. The difference is that the moving forms color is required for the areas that catch the of the motion picture imply more complex types light. Applying physical force to the paper-in of work. Miles away, a generator system creates this example, wrinkling it-creates both power to drive the projector. Electric mains discernible forms and color discontinuities. deliver this power to the projector motor. Film moves through the film gate. Light passes COLOR FIELD MODEL 2 through this film, which is coated with an A more complex model of a color field is emulsion of variable color and density. The provided by a movie screen in a dark room. Its projector bulb must continue to burn, A viewing surface is perceived as a continuous breakdown in any part of this system causes the field of uniform color, if only because any film to stop. The colors that create moving object in the dark appears nominally black. If forms disappear as the screen reverts to its observers were to watch this screen forever, no earlier equilibrium of uniform darkness. change would occur in its color as long as the In the examples of the wrinkled paper and room remained dark. The color is effectively in the movie film, physical work is performed and a state of equilibrium. No physical effort is energy expended to create discontinuities in required to maintain the black color of the what had initially been a field of uniform color. screen, other than that the life support systems Only after the discontinuities are created can we of the observers must keep them alive so they say that forms exist in those respective can continue looking at the screen. two-dimensional fields. Visually perceived Conditions change if a projectionist enters forms can be explained, in the examples and the room and turns on a movie projector. The generally, as color discontinuities in an immediate effect is disruption of the black sur- otherwise uniform field. If a blue disc appears face of the screen. It now appears to consist of in a field of yellow, the disc is visible because a variety of shapes in different colors. When its blueness is color-discontinuous with the this process (the ordinary showing of a movie yellowness of the field. If no discontinuity film) begins, the observers see images of peo- appears and the field remains uniformly yellow, ple, houses, trees, automobiles, and other we may properly conclude no blue disc is at objects.These screen images so closely resem- hand that anyone might report seeing. ble the visual appearance of natural objects that many people cannot distinguish a photograph of a photograph from a photograph of an object. Cinematographers, for this reason, Color as Surface often film live actors playing out their roles The visual world can be imagined as a colored against backgrounds that consist of slides or film or surface. The events occurring among

172 Complementarity in the Visual Field

the colors of that surface result in the images we exists-exists in the sense of possessing see. Although the syntax is unavoidably coordinates in three-dimensional space-dissolve awkward, those events are adequately defined into troubling inquiries into the nature of by assuming that colors "take form." existence. To imagine, instead, that colors are applied Three points are sufficient to locate any to forms would require showing that two- object in three-dimensional space, a familiar dimensional forms are capable of preexisting lesson of high school geometry. Space, for the their colors. This is exactly what cannot be purpose, is imagined as a grid or coordinate demonstrated. No such thing exists as a two- system with intersecting x, y, and z axes. Given dimensional form devoid of color. We see the required coordinates, I can locate myself color even when looking at colorless objects, relative to any other three-dimensional object in because the transparency of these objects the universe. allows the colors of objects behind them to be The lesson has a corollary important to an seen. understanding of color. Unless all three Thus, on the question of whether form or coordinates are available, an object cannot be color comes first, color unambiguously located in three-dimensional space. And the preexists two-dimensional form. To see nothing third coordinate cannot exist for any manifold but turquoise blue, or nothing but yellow, is that is two-dimensional. The two-dimensional possible. Seeing the color of the dark of the space in which images exist is logically night is possible and commonplace. These independent, sufficient unto itself. It cannot be experiences expose us to color fields located in, or by reference to, any system that uninterrupted by forms. They thus give assumes three-dimensionality. evidence of color preexisting form. Art historians and critics have popularized Two-dimensional forms, which are made of the concept of the illusion of space, an illusion colors, are a special category of forms, access- said to appear in the two-dimensional picture ible only through vision. The tactile sense, planes of naturalistic paintings. But visual though it provides information about three- space is not volumetric, because it is two- dimensional space, is blind to the two- dimensional. It cannot be equated with three- dimensional world, to images that appear on dimensional space, if only on the argument, surfaces. This is why touch cannot differenti- drawn from geometry, that an area (in this ate between a one-dollar bill and a twenty- case, the area of a picture plane) cannot be con- dollar bill. Braille books for the blind make use gruent with a volume (in this case, a volume of of dots raised from the surface of the page, not space). because the blind are unable to read ordinary Experience confirms the uniqueness of books by touch, but because nobody is able to two-dimensional universes. I can measure the do so. distance between myself and a table, but the distance between myself and the image I see of a table is as profoundly unreachable as the The Coordinates of Vision beginning of time, hence the argument (which Visual imagery creates the impression of a film misses the point) that the image of a table or surface, a commonplace idea. It inspires the exists only on the retina or only in the brain. equally commonplace questions of where the We might as reasonably argue that three- images are and whether I am only imagining dimensional tables exist only on the fingertips them. Geometry may provide the key to why of the hands that touch them or on the skin questions about whether the visual field of those that bump into them.

Gomplementarity in the Visual Field 173

The difference between two-dimensional ridge or hollow can be found on the surface of and three-dimensional matrices bears on studies the earth that can be identified as the horizon or of binocular vision, or of the unfortunately that a blind person can touch to learn about named "perception" of space and depth. No horizons. We cannot reach the horizon convincing evidence is available that this ourselves, despite its psychological importance purported perception, examined by gestalt and to us. Like the vanishing point resting on it, the perceptual psychologists, is a perception at all horizon continually recedes when approached. in any reasonable sense. It is better explained as Like color, it has no determinable location in an inference from the relative size and location the three-dimensional world. of forms within the field of vision. In the world of thought, the horizon Visual clues to depth in images typically separates day from night. Disorientation in the lack correlation with phenomena in the three- darkness grows directly from the absence of a dimensional world, a characteristic they share horizon or "eye level" that can be used as a with color. In either case, the noncorrelation is focus. Indoors in the daylight, we do not see the evidence of the fundamental two- horizon as a separation of earth and sky. But its dimensionality of the phenomenon. Among a aptly named correlate, the eye level, performs variety of depth clues, overlapping or occlud- the same function. Receding lines on objects ing is the partial "cutting off " of, say, an below my eye level recede upward, receding image of a table by that of a person standing in lines of objects above my eye level recede front of the table. On examination, the actual downward. The complex job of orienting myself table, which has not been cut off in any three- in space depends on reading these signs, seen in dimensional sense, will be discovered the daylight and not in the dark. undamaged. In this and a thousand other cases, Because language and culture encourage we cannot assume that what happens to images the confusion, two-dimensionality is often re- also happens to objects. garded as a fragment of the three-dimensional Perspectival effects are similarly disjoint world, rather than a universe independent from from three-dimensional phenomena. In three- it or parallel to it. The argument subordi- dimensional terms, a mountain is far away if nating the two-dimensional to the three-dimen- walking to it takes a long time. In two- sional proceeds as follows. The top of a table is dimensional terms, the mountain is far away if planar, as is the front of a mirror. Each of these it looks small. Although objects in the distance surfaces, although two-dimensional, has a loca- may look small, this never implies that they tion in three-dimensional space. Its existence at are small, another sign of the dissimilar that location can be confirmed by touch. destinies of images and objects. Vanishing As in the many connections between plane points, lines receding to vanishing points, and and solid geometry, which become less other elements that are constituent parts of convincing when color is introduced into the systems of mechanical drawing or perspective picture, the reasoning has hidden limits. The drawing similarly have no correlates in the only thing that can be confirmed about planar three-dimensional world. The vanishing point surfaces in three-dimensional space is where can be seen but not touched. they are, how shaped (the two-dimensional The horizon on which the vanishing point surface of an egg may be curved), and how lies is meaningful only in visual or two- large relative to other objects in that space. This dimensional terms. Horizons appear in paint- information pertains to the environment around ings as lines of demarcation where the colors the surfaces, not to their interior or essential of the landscape meet those of the sky. But no condition. We cannot confirm, by touch or mea-

174 Complementarity in the Visual Field

surement, what colors the surfaces are or what allow objects to be distinguished from their images, if any, appear on them. The blind, for environments. this reason, cannot identify the images in a The eye must have developed from cells mirror or determine whether a glassy surface is sensitive, as indeed all cells are, to radiant a mirror. energy or its absence: to heat or cold. Muta- The flaw in the confusing literature on depth tion can be imagined in which some cells perception is its insensitivity to relationships became sensitive to that aspect of radiant among visual elements. The visual field is energy we call light and color. The human skin finely tuned to its function. Its elements have to this day has a primitive photosensitivity, more profound purposes than just to help us suggesting that the tactile and the visual can be understand depth. Perspective is often called an traced to a single source. Skin tans in response illusion. I regard it as a magnificent algorithm, a to ultraviolet rays but not to rays of other formula that enables a larger number of wavelengths. elements to be crowded into the visual field Although a surface phenomenon, color is than otherwise would be possible. The visual more complex than the superficial covering it miniaturizing of distant objects solves the was traditionally taken to be. In life forms it problem of how to squeeze more into a limited operates as a code integrated with the genetic box. code. The chlorophylls, for example, have One result is an enormous increase in the been shown to absorb primarily the red and speed with which information can be blue wavelengths of daylight. Familiarly they processed. This increase largely accounts for appear green because they reflect, rather than the superiority of vision to touch. I can look at absorb, light in that wavelength (Govindjee three chairs more quickly than I can investigate and Govindjee 1974). This implies that the cat- then by touch. If the chairs are in the same alyst of photosynthesis is not really light. room, all three can be looked at together. On a Instead, it appears to be the programming of starry night, we see celestial objects too far chlorophyll to utilize some colors of light but away to be touched and at astronomical not others, a programming indicated by chlo- distances from us. The colors of objects allow rophyll's greenness and inseparable from the one object to be distinguished from another and substance.

CHAPTER 20 Color Fields and Colored Forms

In one place [Chevreul] remarks that "red isolated appears differently than when juxtaposed with a white, black, blue, or yellow surface." We should very much like to know how Chevreul isolated his colour. A colour can only be isolated by considering a surface so large that the whole of the retina is occupied, and in this case it can not be used directly for purposes of comparison. George Field, [Field's] Chromatography

f the visual field is to be divided into com- fields. Every visual experience can be identified plementary classes, form and color are as one or the other, and no visual experience unmatched because they lack an ability to consists of both at the same time. I cannot see a be similarly independent. Any color can be black square on a scarlet ground at the same seen in isolation from forms, the experience of time and place I perceive a continuous field of seeing a color field. But two-dimensional forms undifferentiated scarlet. cannot be seen (which means they do not, or cannot, exist) in isolation from color. Rather than functioning as a complement to form, The Union of Form and Field color meets the condition of a visual field An interesting objection arises to a classifica- universal. We may properly assume, as has tion system in which form and field are often been observed, that all visual imagery regarded as alternate modes in which color includes color. makes itself manifest. The objection is that Following after this assumption, a bona fide form must be separable from field to establish complementaritycan be identified in the vis- true complementation. Yet forms are said to ual field. But it is not between form and color. be unable to exist without fields. Objects, in With more defensible logic, the complemen- other words, reside in environments rather tary classes are colored forms, and colored than being surrounded by nothingness.

176 Color Fields and Color Forms

In a two-dimensional matrix, a form is an phenomenological (or topological) terms, no area enclosed by a perimeter; in three form can be, at one and the same time, both dimensions, a volume enclosed by a surface. In inside and outside a perimeter, or both inside either case, no matter how large the form, it is and outside a surface. always possible to imagine fields or spaces An exception exists to the rule that it is lying beyond. A form cannot be removed from always possible to imagine a field larger than surroundings. It can only relocate from one set even the largest form, an environment more of surroundings to another. enormous than the most gigantic object in it. As if the world were a collection of Chinese .The exceptional case is the universe, which boxes, the environmental field surrounding a effectively means interstellar space and the given form may be nested within larger fields. celestial bodies suspended in it. Two puzzling The totality of these nested or concentric fields questions concern its status as object in provides a linkage between the smallest possible environment, if in fact it enjoys such a status. enclosed form, which is finite, and the universal The first is what lies outside, an outside continuum, which presumably cannot be apparently as inaccessible as the back of a exceeded. In a simple example of the reflection in a mirror. The second is whether the mechanism of environment, which functions in universe, whether in steady state or expanding, both three dimensions and two, I am located in could be uniquely an object with no my room. My room is in, respectively, my environment surrounding it. house, the city, the country, the earth, the If the universe, from microcosm to universe, and any enclosures that exist beyond macrocosm, is a continuum of nested that. Because a field is environmental to environmental fields, there are only three final anything and everything within it, the earth is possibilities. Each requires the concept of the environment for my city and not merely for infinity. There may be an infinite number of me. these fields (each enclosed in a field larger than The night sky is imagined as a field without itself ). Or the last field may be unenclased, edges when human beings wonder if its and therefore infinite in extent. Or both blackness continues forever. The stars are conditions may be true at the same time. forms lying within it. The primary difference The modern conception of the universe as between forms and fields, alternately called finite implies a finite number of fields, with the figures and grounds, is that fields need not be last field finite as well. But it appears to be enclosed. The unenclosed field, which is merely a variation of the second case, and the limitless, is evidently the paradigm for all concept of infinity has been displaced rather fields. There is no other way of explaining why than genuinely eliminated. What is infinite in geometry books, which give instructions about an enclosed, finite universe with nothing how to construct various enclosed forms, offer outside is the exterior "nothing," unbounded no corresponding recipes for constructing edges unless a way can be found to identify its limits. of fields. Perhaps it is tacitly recognized that Children sometimes ask where the end of fields with edges are actually forms. the world is and what ties beyond. The Any entity with a finite boundary questions are easily expanded into others for (archetypally, a form) retains a potential for which adults similarly have no answers. How enclosure within that which is infinite in extent can we identify nothing when we see it, if it is (archetypally, a field). That the converse is not something we cannot see? If it cannot be seen true is merely a way of saying that the word or identified, how could a determination be infinite means uncontainable by the finite. In made about whether nothing surrounds the uni-

Color Fields and Color Forms 177 verse? Even a mathematical conception of the field rather than form within human perception. nature of the universe, which led, for example, Irrespective of whether the darkness has a to the idea that it was finite but had no perimeter, the human observer confined to a environment, could not eliminate the question fixed place lacks the capacity to see it or to see of how this nonenvironmental nothing could be beyond it. identified or defined. What is implied is not Form and field, as this suggests, are merely an absence of subatomic particles but an essentially contextual. A form in one absence, as well, of the empty spaces between environment may function as an environment them, the nothing of the universe in which we for another form enclosed within itself. The live. question of how to separate form and field, if The inconsistency in the reasoning is that we they can be separated, turns essentially on are apparently asked to assume the existence of location. Or it turns on location respective to a two distinctly different types of nothing, unrelated to one another. The first lies within the universe and is variously identified with Separating Form and Field either empty volumetric space or the color black A black disc against a gray background in a (as in the perceived blackness of the spaces two-dimensional matrix provides a simple between the stars). The second nothing, outside example of a form in a field, a configuration the universe, is among other things an absence psychologists often refer to as a figure on a of the first, because the conception of a universe ground. Whether the figure is regarded as on with nothing outside does not allow for exterior its ground or in it varies, but it is a question spaces. The second nothing presumably cannot of topological interest. Kofflca preferred on: look black. If it did, we would be compelled to "The left cross can appear either as a blue cross imagine a universe afloat in a sea of that on a yellow ground, or as a yellow cross on particular color. a blue ground .... The ground is unaffected These and similar questions are applicable by the contour and is partly hidden by the to worlds of two dimensions, rather than figure, yet it lies without interruption behind the unique to those of three. It might even be figure .... No visual figure can occur without a argued that their origin lies in the attempt to ground on which it appears" (Koffka 1922, understand perceptual experience at the edge 551). Kohler preferred in or within: "The of the visual field, a frontier that similarly threshold for a patch of color has been found to defies efforts to reach beyond it. What I see at be higher in the area of a figure than within a any given moment is finite but never all there ground of the same objective color" (Kohler is to see. Like the images in mirrors, what 1947, 120). remains to be seen apparently has no end. That any figure might lie on its ground is Metaphorically, I might want to imagine that it a dubious assumption, because the word reaches to the end of the world. implies superimposed on. In a two- The perimeters that define two-dimensional dimensional universe, nothing can consistently forms also provide the boundaries at which be assumed to be superimposed on anything form and field visually meet. For example, I else. Koffka's further assumption that a ground might imagine exhibiting a blue disc so large (or field) might continue "without interrup- you would be unable to see either its edges or tion" behind a figure (or form) is as monumen- the environment surrounding it. The disc, for tally unverifiable. A black cross in a gray field you, would be a field, not a form. On similar or ground cannot be lifted to investigate reasoning, the darkness of a starless night is whether the gray continues behind it ( figure

178 Color Fields and Color Forms

implicitly three-dimensional. The gray field passes, in each, behind the black cross, an impossibility in vision because the images we see have no depth. The subtleties of on versus in may seem abstruse. But most people have an excellent practical understanding of the structure of two- dimensional fields. Imagine, for example, that the black cross in gray field was to be enlarged, to be painted on the side of a building. It

Figure 20-1. A black figure on a gray ground. If superimposition of one form "on" another cannot be assumed in a two-dimensional universe, figures (forms) must lie "in" grounds (fields), rather than on them.

20-1). There is no "behind" in a planar uni- verse, because any term that relies on the con- cept of a third dimension is experientially meaningless when that dimension is absent. The question of whether a given human observer exhibits sensibility to visual phenomena reduces in large part to the ques- tion of whether that point is understood or denied. A model of the black cross and gray ground can be constructed by cutting the shapes out of colored paper. Figure 20-2 illustrates three methods. Although indistinguishable from one another when completed, the gray field is con- structed differently in each. In the first two, the ground continues, as Koffka imagines, behind the figure. In the third it does not. In the second model, a square hole occurs in that part of the field concealed by the black cross. The third is the only model that accurately represents either a two-dimensional universe Figure 20-2. Three models of a figure in a ground. The three models of a black cross on a gray ground or the manner in which form and color are look visually similar but differ in construction. Only seen in the visual field. The other models are C is genuinely two-dimensional.

Color Fields and Color Forms 179 would be necessary to calculate how many tains it, and the two forms share a common set gallons of gray paint would be needed. Even of edges. The exterior edges of the black cross children, given the problem, are able to are the interior edges of the hole in the center of understand that the final visual effect will be the gray shape. the same whether or not extra paint is used to This suggests that fields (by which I mean continue the gray area "behind" the black cross. uninterrupted fields) are unable to coexist with The practical solution is as in figure 20-2(C). forms. Colored forms and colored fields are Gray paint is not needed for that portion of the bona fide complements within the confines of field occupied by the black cross. the visual field, although form and color are not. The computations for the gray paint point to Anything that can be seen is either color- a method for separating form from field (or continuous or color-discontinuous. It is either a figure from ground). The gray area of figure colored field or a collection of at least two 20-2(C), although likely to be regarded as a colored forms. continuous field (passing, perhaps, even behind the figure), is actually discontinuous. The discontinuity consists of the hole in its center, Noninterpenetrability corresponding to the space occupied by the No contemporary understanding of opposite- black cross. The gray shape is only nominally a ness explains Aristotle's assertion that " ‘grey’ field, more correctly a form. Like any form, it and `white' do not apply at the same time to the consists of a fractional part of a field, and has a same thing, and hence their constituents are perimeter or limit. Its perimeter is identical opposite" (Metaphysics, book 5). It is true, with that of the black cross. however, that two colors cannot occupy the As this suggests, the type of configuration same visual space at the same time. Colors are misleadingly classified as a form in a field (or noninterpenetrable. Among conditions that a figure on a ground) is more coherently follow from this, no entity, including a understood as two forms. One is the negative perceived color spot, can be both entirely red shape, or complement, to the other. Although and entirely blue at the same time. Euclidean geometry, oblivious to negative The noninterpenetrability of colors sug- space, fails to make clear the dynamics, forms gests an attribute often assumed to be exclu- can only be seen two or more at a time, sive to material objects. Just as two colors because every form effects disruption of a cannot occupy the same visual space at the field. In that truism familiar in the visual arts same time, two corporeal objects cannot occupy but often overlooked in studies of perception, the same physical space at the same time. every form has a negative shape to itself. And Colors, however, are devoid of mass, the the figure on ground is actually two figures, quality likely to be cited to explain the inability another way of expressing the idea that objects of three-dimensional objects to interpenetrate. possess environments. The mass of a corporeal object, it seems Colored forms can be defined, in visual reasonable to believe, is what cannot move into terms,as fractions or portions of colored fields. the space occupied by the mass of some other What I see at any given moment, if blue, is object. either all blue (a colored field) or partly blue This raises the question of what we mean (a colored form). In the illustration, the entire by saying that objects have mass and colors do configuration consists of two forms. One is not. Beyond that, I am interested in whether gray; the other, black. Each is a fraction of a mass is really necessary to noninterpenetrabil- field, with edges within the domain that con- ity, A more likely answer might be surface.

180 Color Fields and Color Forms

Even a child learns to understand the constitutes that which is delimited in the other. difference between colors and objects and, Because noninterpenetrability and surface are indeed, must do so to learn how to use color common to both two- and three-dimensional names. Chairs, apples, and airplanes are universes, it seems reasonable to ask what the accessible to multiple senses; they can be both relationship is between them. seen and touched. Colors are available to a single sense. They can be understood only through visual means. If mass is an essential Displacement difference between a color and an apple, the That familiar catechism of elementary school word appears to apply exclusively to extension science class, that two objects cannot occupy within a third dimension; to the type of the same space at the same time, suggests that extension that can be touched. what is being attributed to corporeal objects is I am going to move from this to propose that as likely to be a property of space. Apparently whereas mass is associated with extension into a nothing can move into a space that is not third dimension, noninterpenetrability is as empty, much as nothing can be put in a box that intimately associated with two-dimensionality. I is full. An occupied space becomes available include among two-dimensional entities not for use only if its contents are moved or only planar universes (for example, the visual displaced. In a familiar example used to field or the picture plane) but also the shaped illustrate the nature of displacement in the and colored two-dimensional surfaces of three-dimensional world, I settle into my three-dimensional objects. Each of these bathtub, causing the water level to rise. surfaces can be defined as a twodimensional Following a principle said to have been plane meeting itself in every direction, the enunciated by Archimedes in his own bath, the characteristic that enables it to enclose what we water has been displaced by me. Because identify as the volume of the object. Its matter is noninterpenetrable, two objects cannot analogue in the two-dimensional world is occupy the same space at the same time. perimeter, a line that meets itself to enclose an Color displacements that occur in the visual area rather than a volume. field are similar to displacements in the Surface is the two-dimensional (planar) physical world in that they similarly serve to aspect of any three-dimensional form. Perim- sustain the constancy of the field. But they eter is the one-dimensional (linear) aspect of cannot be explained by a similar analogy, in any two-dimensional shape. The hierarchy can part because the visual field lacks empty spaces be carried further by noting that length is the to which displaced colors might migrate. As a null-dimensional aspect of any line, null- simple example of visual field displacement, I dimensional because it does not include width. might imagine observing a field of yellow into Surface, perimeter, and length are absolute which an orange disc has been introduced. delimiters. I mean by this that any three- Because I cannot see yellow where I see dimensional object can be defined as the orange, the orange color of the disc displaces a volume enclosed by its surface; any two- portion of the yellow. In operational terms, dimensional form, as the surface enclosed by while looking at the orange I cannot see the its perimeter. Any line can be defined in terms portion of yellow nominally behind it. The area of its length, which in fact is how we define lost to view, or displaced, is identical in size lines. Surface, as this further suggests, is and shape to the occluding disc. common to both two-dimensional and three- It is not possible to say where the occluded dimensional forms. It delimits in one case and color has gone. It cannot be genuinely behind

Color Fields and Color Forms 181 the disc, a meaningless term in a universe with energy in the universe, for example, is said to only height and width. Unlike the water in be constant. We have no idea whether this is Archimedes' bathtub, it has not moved else- also true with regard to the amount of color. where to displace some other item. It has Imagine, to illustrate the nature of the question, merely been displaced from the field of vision, that a green table burns to black ash. Although vanished into a world of interphenomena. It the pile of ashes is smaller than the table, the was visible once; in some cases it may appear chemist is prepared to account for the again. But it is not there (or anywhere) now, in permutations of every molecule, to prove that any perceivable or conceivable sense. There is nothing was really gained or lost. no answer to the question of where what we No similarly exhaustive audit can be see goes when for any reason it is no longer provided for the metamorphoses of color. seen. Where did the green of the table go? Where did The primary difference between visual dis- the black of the ashes come from? Insofar as placement (of colors) and physical displace- motion is implied in either case-and it is ment (of matter) is that one is a series of virtually impossible to describe the phenome- transformations; the other, a series of motions. non without that metaphor-the journeying of Displacement, in the physical world, resembles the greenness of the table differs from the a game of musical chairs. The path of each dis- travels of its atoms. Unlike motion in the cor- placed item can be traced, to show it has poreal world, it is not a journey from here to merely moved on to displace something else. there, because there is remarkably elusive. Like At least ideally, when I step into the filled light, which similarly "travels" without neces- bathtub, my body displaces water, which in sarily arriving any place in particular, colors rising displaces air. That air moves to fill the can apparently move on to an ambiguous des- void I had occupied before getting into the tub. tination. Or, like the blackness of the ashes that The bathtub model has limits. It might be we formerly knew as the greenness of the table, construed to incorrectly imply, for example, they can appear as if by spontaneous that no order determines what displaces what. generation. Yet I cannot sink into my bathtub if its water The question of why noninterpenetrability is frozen into a block of ice. And if an ordinary ought to exist at all comes easily to mind. Solid drinking glass is pushed, bottom up, into objects are said to actually consist primarily of water, the water will be unable to displace the empty space in which subatomic particles float air in the glass. What displaces what, and at relatively enormous distances from one under what circumstances, is contextual. It is another. If the table we see and touch is not also associated with surface, the attribute genuinely solid, why does it persist in behaving common to both three-dimensional objects and as if it were? Why cannot the cloud of color. What allows my body to displace water molecules that constitutes one object pass, more easily than ice is that water has a less ghostlike, into, or through, the cloud of resisting surface than ice. molecules that constitutes another? A common explanation compares the motion of molecules in an object to the action Constancy of Color of a moving airplane propeller, effectively bar- Many questions that have been regarded as sig- ring access to the space in which the move- nificant when asked about universes with three ment occurs. This explanation cannot be dimensions are rarely or never asked about extended to colors (there are no molecules on those that have two. The amount of matter and which to hang the propellers), which suggests

182 Color Fields and Color Forms

the analogy is weak. But it is amusing enough until the tension is broken. Soap facilitates the to pursue to a point. If the molecules of an mixing of oil and water by weakening this object are imagined to be equipped with the tension. hypothetical propellers or comparable Surfaces constitute a large part of the mechanisms, these devices would not neces- integrity of any object. By surface I mean the sarily be required throughout. They could most actual two-dimensional shaped plane, efficiently be attached solely to the molecules irrespective of what lies beneath it. Life forms on the surface of the object. That surface, which resemble complex combinations of liquids and incidentally displays the object's color, is what jellies rather than gases or genuinely dry solids. most immediately encounters the surfaces of Most or all possess specialized surfaces other objects, offering more or less resistance to designed to protect their interiors from, among intrusion. other things, dehydration. The loss of a large amount of skin is fatal to higher organisms, just as destruction of the cell wall is fatal to an Color as Surface amoeba. Human epidermis, an example of Unlike the surface of an onion, that of an image surface in a life form, has an outside (the actual cannot be peeled away. This evidently accounts surface) and an inside. It cannot function for the tenacity of color in the visual field or for correctly if grafted to a human body so that its its remarkable constancy. I cannot add to, or outside is inside. subtract from, the totality of what I see, the Although the surface tension of liquids can exact reason I cannot see both red and blue at be broken, that of solids displays a remarkable the same time and place. integrity. In a homogeneous substance, it One reason for suspecting that displacement typically cannot be weakened, divided, or may be a function of surface, consequently of destroyed. If a piece of stone is broken into two-dimensionality, is its more radical nature in smaller pieces, each fragment has as hard a the visual world. Displaced color spots are surface as the original. The homogeneity is obliterated, rather than being transferred from similar to that of a color field, in which, for one place to another. Furthermore, resistance to instance, any portion of a field of displacement in the threedimensional world can homogeneous green is as green as the whole. be correlated with the character of an object's Whether because of the nature of surface more easily than with other physical two-dimensionality (surface) or for some other aspects of the object. reason, extension, for either objects or colors, In the continuum of resistances, gases, appears to imply extension of the capacity for which have no surface, present no barrier to being noninterpenetrable. mingling with one another. Although an atom of It would be regarded as a retreat into solip- chlorine presumably cannot occupy the sism to assume that the world did not exist spatiotemporal location of an atom of oxygen, other than in the form of the colored images I this does not, as a matter of fact, prevent a cloud see. But perhaps this trivializes the issue. The of chlorine gas from mixing with a cloud of visual field is a surface, a planar matrix com- oxygen, evidence that three-dimensionality per posed of color which, as such, has certain se is not a barrier to penetrability except at the characteristics. Beyond this, the entire world molecular level. On a planet so hot that all is a surface, including those aspects accessible elements were gaseous, all could presum- to touch. A hologram looks more convincingly ably intermingle. Liquids display what is called three-dimensional than an ordinary photo- surface tension, often resisting other liquids graph, apparently solely because it more fully

Color Fields and Color Forms 183 records the nature of the surfaces of objects. ing on the artificial lines we invent in making The external world, as has been said of the outline drawings of objects. The circle I draw picture plane of a painting, is a surface that to depict the moon is not the moon's perime- shuts us out, compelling each of us to remain in ter. It is, however, the perimeter of that por- his or her assigned place. The essence or tion of the moon's surface I am able to see a the interior of any object remains perennially moment, an essentially two-dimensional inaccessible. The limit is only transcended in configuration. dreams, fantasies, and myth. Ghosts, Perimeters and surfaces are inversely related apparitions, and spirits, as we imagine them or to the quantity of what they enclose, a depict them in movies, pass into or through relationship that cannot be altered. Cutting a solid objects, an expression of the idea that square piece of paper in half does not reduce the other worlds may not be subject to the total area of the two pieces; but their aggregate limitations of this one. Lovers, imitating perimeter is increased by 50 percent. Smashing Hermes and Aphrodite, may imagine merging a brick into fragments does not affect total with one another. Mystics can yearn for union volume or weight. But the total amount of with God, in whatever form they imagine that surface area in all the fragments vastly exceeds union. that of the original brick. The figures of ghosts and ghostlike objects, Because surface and perimeter are basic probably invented after looking at clouds and physical functions of the forms they delimit, we fogs, suggest ways of imagining what it might cannot alter them without changing the forms. be like if two objects were able to occupy the Peeling the outer layers from an onion does not same space at the same time. But nothing helps create an onion without a surface. It produces in the case of colors. I cannot even imagine only a somewhat reduced onion with a smaller, what it would be like if something I saw were but entire, surface of its own. entirely blue but also entirely green at the same time. The limit is not genuinely transcended by imagining the colors mixed together. Perimeters Forms and fields can be visually perceived if, Geometry of Surface and only if, composed of color. Forms have in an alternate geometry based on vision, any outer boundaries or perimeters, which cannot be two-dimensional form, including a perceived true of fields if they extend to infinity. Because color spot, might be defined as the area the enclosed forms of plane geometry are enclosed by its perimeter; any three- described by telling how to construct their dimensional form, as the volume enclosed by perimeters, an alternate to imagining them as its surface. No object can be both two- shapes is to envision them as locations. dimensional and three-dimensional at the same A triangle, for example, is any enclosed area time, which essentially means that enclosed located within a perimeter composed of three volumetric forms cannot maintain linear straight lines. The area outside a triangle's perimeters. Imagine, to illustrate the perimeter (the expanse from the perimeter to mechanism, that a black bowling ball is painted infinity) is not called a triangle. But it is limited green. The small area of green paint used to at its interior edge by the perimeter it shares begin the task has an identifiable perimeter. with the triangle. Axioms for a two- But when the entire surface has been covered dimensional universe might be as follows: with green paint, the perimeter of the green 1. A perimeter is a line that returns to its area disappears. This phenomenon has a bear starting point.

184 Color Fields and Color Forms

2. Perimeters cannot be seen unless the because nothing can be located both inside areas they separate differ in color. and outside of a given perimeter at the same 3. Forms are areas marked off by perim- time. eters. 6. Because nested, or concentric, perimeters 4. Fields are areas devoid of perimeters. are possible, a figure that lies in a ground can 5. No positive shape can be its own nega- serve as a ground for another figure lying tive shape (or no figure can be its own ground), within itself.

PART FOUR Color and Culture

I am persuaded . . . that no two colours produce together a positively unpleasant effect. William Benson, Principles of the Science of Colour Concisely Stated to Aid and Promote Their Useful Application in The Decorative Arts

CHAPTER 21 Hue, Color, and Culture

Hue can be described as the property which gives color its name-blue, blue-green, red, etc.-the name, that is, by which it is distinguished from other colors in the . J. Scott Taylor, A Simple Explanation of the

hildren in the lower grades are taught of ideal forms palely reflected in the forms of to memorize the word sequence "red, everyday experience. The question turns back orange, yellow, green, blue, violet" on itself, dissolving into nonsense. The English but not because a special value exists in language offers no syntactical barrier to placing knowing the order of the colors of the the words pure or ideal before the name of any spectrum. The names also identify the hues, color. A world of ideals, if it exists, might seem the major varieties of color other than black likely to include pure blue, as well as pure (or and white. The use of a single set of names for ideal) varieties of other pure or impure colors. both purposes reflects the belief that the hues The inconsistency is that blue is a perceptual are found in their purest or most typical form experience, and ideals are usually thought to be in the spectrum. A person asked what pure beyond perceiving. Dreaming of ideal blue looks like can point to the blue portion of perceptions that can never be perceived, ideal the spectrum. Spectral blue is regarded as experiences that can never be experienced, either pure blue or its closest approximation. becomes self-contradictory. Can a blue that is more pure or more Nevertheless, pure, when applied to color, intense than the blue of the spectrum exist? is a popular descriptive term. Almost everyone The question asks whether we can imagine understands the implied antithesis between such a blue, perhaps located in Plato's world pure and impure or between perfect and

187

188 Hue, Color, and Culture

imperfect. In descriptions of color, as all other colors from its location. I conclude elsewhere, pure and perfect pass as ad hoc that colors not included in the spectrum, say, synonyms for beautiful. Sometimes pure is an silver, gold, black, white, and the iridescent alternate for bright. Rarely is the word applied colors, are not substantively different from to colors that are not spectral hues or close to red, blue, or green. They differ in color but the hues. We do not speak of, say, pure olive not in whether they are colors. For experien- drab, pure medium gray, pure brown, or pure tial as well as conceptual reasons, we cannot navy blue. have a category of colors that are not really Before acquiring its additional meanings, colors. purity was a traditional parameter for grading Colors can be regarded as objects of per- color. Its less judgmental modern descendant, ception, thus objects. They are more often chromaticity, means saturation (or proportion) called properties of other entities to which of hue. As chromaticity or hue content drops, they are said to belong. Dictionaries often colors become, in the vernacular, paler, darker, report, say, that red is a property of one por- duller, or muddier. tion of the solar spectrum. This means that Children encouraged to recite the names of both red and a particular portion of the the hues are usually encountering their first, spectrum can be seen at the same time and and unless they study art, their last, formal place. By this ill-considered criterion, red can instruction about color. Students and teachers be linked to anything that is red. Red is the are likely to describe the exercise as learning color of blood, of red flowers, of red laser the names of the colors rather than those of beams. That red and the other hues are found in the spectral hues. The blurring of distinction the solar spectrum is scarcely more useful than perpetuates the confusing practice of using the information that penguins, giraffes, and color to mean, ambiguously, either all colors elephants are found in zoos. It tells us where, or only the hues. The looseness of language has from time to time, the hues are located, but not an acrimonious history. Goethe criticized what they are. Newton for devising a theory of color based The word red can be defined in a less con- only on the colors of the solar spectrum. The fusing way. We can call it a name convention- limitation might be defended if the spectral ally conferred on a portion of the spectrum, hues could be shown to enjoy special status, or rather than a quality belonging to the spectrum if colors absent from the spectrum are not really or to some part of it. This eliminates any need colors. to explain the nature of a quality or to justify An otherwise inexplicable body of formal drawing a distinction between a quality (or and informal argumentation addresses exactly property) and an object. Although the object- this issue. We are told that, for example, black, or-quality conundrum has persisted through- white, gray, metallic colors, fluorescent colors, out the history of color theory, the differences and iridescent colors are not really colors in the between objects and qualities relate to human same sense as the spectral hues and their value systems rather than to the phenomeno- derivatives. The argument proposes a class of logical world. items that effectively are not really colors. Yet Reasoning about the nature of color and each of the not really colors is a color by the light is often argumentation about their place criterion of exclusion. in a hierarchy, a subjective concern.' White Assume that a color is an entity endowed light can be divided into colored rays, which with spatiotemporal location in the visual field. Newton offered as evidence that light is not We learn, by seeing, that this color excludes a quality. Light enjoys, Newton argued, a

Hue, Color, and Culture 189 higher status than qualityship can confer. His separated from visual phenomenology. No curious logic is that no quality possesses theory of forms is sufficient to explain the qualities. Yet color is said to have the qualities difference between day and night, a difference of hue, value, and chroma. Color can be, say, that is the locus for our ideas about both light moderately dark, very light, bright, dull, or and time. glossy, all similarly called qualities. We would Newton may have meant to elevate light to have less need for adverbs if qualities could not the status of virtual object, a way of saying he be further qualified. Every quality possesses thought light was important. At earlier dates, qualities, including the quality of being present other people said essentially the same thing. or not present. During the twelfth century, Abbot Suger Newton's purpose is more interesting than thought light was a symbol of God or an his argument. He wanted to take issue with the emblem of God's works. The association assumption that material objects are of special appears in the Bible. Ancient peoples were significance, a significance we attribute to probably more impressed by the effects worked them by calling them real. Light, color, empty by light than by any attributable to the surface space, and other immaterial entities are less colors of objects. Light causes growth. In significant, not entirely real, an idea with excess, it destroys. It can be concentrated ancient roots. For Plato, geometrical forms through lenses to start fires. We know today were worth discussing. The empty spaces that objects are not passive reflectors. Plants around them were not. A more modern grow not just because of the light of the sun but conception of space pictures it as a system of because their green pigment, chlorophyll, coordinates, never wholly empty because it reflects some rays while absorbing others. And includes these coordinates. But any coordinate the dangerous aspects of ordinary sunlight are system depends on an arbitrarily selected matched by those of a ruby laser, a center from which its axes originate. No fixed concentration of one wavelength or color to the center is available in the universe, therefore no exclusion of the rest. absolute center for a universal coordinate For ancient peoples, the rainbow could be system. The modern idea that no coordinate construed to provide a reason for according a system is privileged is close to Aristotle's special status to the hues. It comes and goes in argument that location is an artificial concept. I the sky without human intervention, suggesting may well be convinced that a chair is there. But colors belonging to God or the gods. When it there effectively means at an indeterminate appears in myth, as in the Bible, we find it place. It remains as mythic as once upon a accorded a special place. In myths of India, time, which means at an indeterminate time. Indra carries a thunderbolt, like the Greek god The looseness of there has a bearing on Zeus, but also the rainbow. Part of a creation arguments meant to prove that chairs and other myth says the gods and titans churned a material objects are there but colors are not. primeval ocean of milk. Airavata, a sacred The ambiguity of location or the arbitrari- milk-white elephant whose name means ness of coordinate systems is not the whole rainbow, was one of the first creatures born story. Light passes any designated point in from the churned milk (Zimmer 1946, 104). space with a rapidity once thought to be evi- The idea that the rainbow and its hues have a dence of instantaneous propagation. Because special significance, feeding on itself over the it never stops moving, light cannot be said to centuries, made the hues easy to overestimate. reside at a certain fixed distance from some For Goethe, the German language "has the other object. Neither, however, can light be advantage of possessing four monosyllabic

190 Hue, Color, and Culture

names no longer to be traced to their origin, viz., White is called "not a color," though also, gelb, blau, rot, grun (yellow, blue, red, green)" "light of all colors," or even "every color." (Matthaei 1971, 249). Goethe was too Each of these propositions is dubious, and they sophisticated to argue that the hues were more make no sense as a group. Newton's prism important than other colors because God had experiments justify the assertion that white put them in the sky in the rainbow. But he light is composed of rays of the spectral colors. assumed it would make sense if the names of But the rays in aggregate are not, as Newton hues had always been constant in meaning. called them, "all colors." The metallic colors, Red is not blue, or the name blue is used for example, are not matched by spectral rays incorrectly if applied to colors that ought to of those colors. White (light) is not be called red. The group of colors that synonymous with white (the color), and speakers of English call red can be designated generalizations about the whiteness of light do by any other name, and the conventions for not hold for the whiteness of other objects. segregating colors can be modified. Red can be White paint is not composed of paint of all defined as a larger class, a smaller class, or colors, any more than white shoes are eliminated entirely. Its relationship to other composed of shoes of all colors. colors, including white, can be regarded Does white light separate into colored rays differently. because it is white or because it is light? Light, Among speakers of English, pink is consid- not white, is what passes through prisms, the ered a variety of red. The name is applied to only known instance in which anything white colors that can be matched by mixing red paint can be shown to be composed of colors. That and white paint, especially if a high proportion white is not a color is as faulty as the of white is needed. Intelligent beings from assumption that white is all colors, except to another galaxy might argue that pink should the limited extent that white is not a spectral not be classified as a type of red, because pinks hue. I cannot see white, just as I cannot see look different from reds. The argument is yellow, at the same time and place I see blue. irrefutable. The visual difference is the very White is regarded as pure in a different sense reason we segregate certain pale reds by call- than the spectral colors because it is also called ing them pinks. colorless. James McNeil Whistler responded Red is called pure because a hue. White is angrily to the art critic P. G. Hamerton, who called pure for other reasons. Pink is not complained that Whistler's Sym phony in White usually called pure, though made from two No. 111 was not an entirely white painting: "pure" colors. What rules do we follow in "Bon Dieu! did this wise person expect white deciding a color is pink, and in deciding hair and chalked faces?" (Whistler 1890, 45). whether pink is pure? Any pink can be repli- Hamerton was insensitive but not incorrect, and cated by a mixture containing some red, his readers probably understood him. Speakers though this is also true of any purple or any of English are rarely challenged for the orange. Purples and oranges are neither called assumption that white, because pure or not a pink nor categorized as types of red. I con- color, ought to be wholly untinged by other clude the red component of pink is not the colors. issue, though pink is called a type of red. The The reasoning carries over into racial desig- critical factor is our often confusing concep- nation, an elaborate code that only nominally tion of white, matching our confusing concep- refers to skin color. Nobody's skin is actually tion of black. black, white, yellow, or red. Because white

Hue, Color, and Culture 191 and black also are used as figures for good and at 6:30 P.M., Sept. 12, 1925. The present bad, the English language is heavily endowed writer (X) and another observer (Y), with idioms that can be mistaken for (or turned whose appreciation of color values is very into) racial slurs. Lily white intentions are pure; accurate, independently recorded the black motives are evil. colors observed, in order from the top to The idea that white is, or ought to be, pure the bottom (i.e., theoretically from red to or untinged by any other color is often violet) of the uppermost or clearest arc. expressed by calling it colorless, an These observations are recorded below. illconsidered usage because "colorless" is best The colors in italics are those that seemed reserved for designating transparency. Between to stand out most plainly. As the bow was water and milk, we all know which is white in visible so long, these observations are the color and which is colorless. Yet for result of careful judgment. Their disparity Greenough, "the absence of color in the teeth is shows: 1. the variation in what is seen by as beautiful as its presence in the lips" the eyes of different individuals and in the (Greenough 1962, 90). White teeth are neither interpretation their minds make of what colorless (transparent) nor the same color they see; 2. that the rainbow does not among different individuals. Indeed, matching consist of seven ordered bands or distinct a cap to an adult's teeth usually means finding colors; 3. that the ancient descriptions of the right shade of ivory, and dentists routinely it as red, yellow-green and [purple] were use charts of human tooth colors. nearer what one actually sees, than are the Greenough's remark implies that white, spectral colors. because not a color, does not count. The wide currency of a similar belief must explain why any pink is more likely to be thought of as a type of red than a type of white, even if its white component exceeds the red: the white is not counted as a color. Or, as in the politics of race, white is used as if the word meant entirely white, a standard not applied to other colors. Hue weighs more heavily than chroma or value in our assessment of color but appears to be arbitrary.The classical Greeks named the colors of the rainbow as red,yellow-green, and purple (Wallace 1927, 24), a close approxima- tion for how the colors appear in modern photographs of spectra produced by diffraction gratings (Hoyle 1957, 183). Wallace argued, on evidence of a rainbow she had seen, that In Islamic cosmology, the colors of the the Greek naming of its colors was more spectrum are identified as "red, yellow, green, accurate than our own (Wallace 1927, 76): and blue, corresponding to the four elements" and linked to celestial bodies (Nasr 1978, A magnificent double rainbow that lasted 86-91). Mars is associated with red, the sun About half an hour was seen in London, with yellow, Jupiter with green, and Venus

192 Hue, Color, and Culture

with blue. Saturn and the moon are respec- guished from the other. At a distance, or under tively linked with black and white. poor illumination, navy blue can be confused Modern ways of dividing the spectrum are with dark green more easily than with pale as arbitrary as those of prescientific peoples. blue. I conclude that navy blue does not look Newton, who expected the objects of the like pale blue as much as it looks like dark world to be ordered in sets of seven, saw seven green. colors in the solar spectrum. In England the The criterion for calling colors blue is not colors are still named as he saw them (indigo is that they took alike but that they share a com- retained as a major hue), probably because mon hue component. Bias in favor of the hue Newton was English. In the United States, parameter is built into everyday forms. If color wheels usually include six hues, perhaps asked to describe a dark blue color, the in deference to American practicality. As we respondent is thought to be correct if the color learn in high school geometry, a circle can be is identified as blue. It cannot be called merely divided into six equal sectors by marking off dark, a term that is not a bona fide color name. the radius on the circumference. Only a com- But dark blue is dark as much as it is blue. pass is needed. A protractor is required to Conventions in color naming, like other divide a circle into seven sectors. Furthermore, conventions, are not necessarily consistent seven is an odd number, which complicates the among themselves. Pink is a special name for task of dividing colors into pairs of com- light reds, , a less frequently used plements. name for light blues. Light green is sometimes Divisions within hue classes are set for simi- called mint green. Light yellows are larly eccentric reasons. Perceptible differences sometimes called cream color, and two inconsistent with preferred sorting criteria are observers may disagree about whether cream is ignored. By a visual standard, a bright red may a type of light yellow or a type of yellowish resemble a bright orange more than a dark white. Brown and gray are enormous color maroon resembles a pale pink. Yet the first classes that cut across the boundaries of hue. two, because hue is the preferred criterion, Some browns tend toward orange, others are will be said to belong to different color classes. yellowish, greenish, or reddish. But all are The last two will be assigned to the same class, cataloged as browns. Pale brown is usually because both are varieties of red. A visual called tan or . Brown with a high blue standard for color similarity is the simplest of component may be called gray. A wide range all standards. We need to determine from look- of colors intermediate between brown and gray ing at the colors how easily one can be distin- has no particular name.

CHAPTER 22 Prime Minister Gladstone and the Blues

By that same way the direfull dames doe drive Their mournful charett, fil'd with rusty blood. Edmund Spenser, The Faerie Queene

othing is phonemically, morphemic- conventional), no common understanding can ally, or expressively significant about occur among the peoples of the world about the group of sounds that form the what the terms similar and dissimilar mean spoken word red or the group of letters by when applied to color. which the name is written. In German, rot and In the social sciences, Segall, Campbell, and Rot (respectively, the adjective and noun forms) Herskovits reported that the topic of " `racial' are used. In Spanish, red is rojo; in French, or cultural differences in color perception and rouge; in Dutch, rood; in Italian, roso; in Polish, in color vocabulary, has the longest and most czerwony. sustained research history in the culture and No matter the language, the color class red perception area" (Segall, Campbell, and Her- is thought self-evident. A range of colors skovits 1966, 38). The grandfather of the sub- belongs in the class, the individual shades in stantial literature is William Ewart Gladstone this range look similar according to a univer- (1809-98), Queen Victoria's prime minister. sally acknowledged sorting criterion. Only the Gladstone was an amateur philologist and stu- name of the class (its spelling and pronuncia- dent of the classics. A study of the Iliad and tion) would be expected to differ from one Odyssey led him to believe that Homer (ca. language to another. If, however, no universal 800 s.c.) applied color names inappropriately. sorting criterion exists (if sorting by hue is just Citing such phrases as "black blood," Glad-

194 Prime Minister Gladstone and the Blues

stone argued that the usage could only be common in anthropological circles, they have understood by assuming that Homer, defective acuity for blue (possibly also green), traditionally said to have been blind, was which they confuse with black. instead colorblind. Consequently, Homer Gladstone and his followers took counts of thought blood was black. The poet's the number of times color names are used in contemporaries are not known to have ancient and modern literary works. Attention questioned the usage, inspiring Gladstone to was paid to usage that seemed uncommon. argue, further, that the ancient Greeks as an Employing the same name for more that one entire people had little or no ability to perceive color was taken to be a sign of inability to color (Gladstone 1858, 3:457- 99; 1877). distinguish between the colors. In Beowulf, If Gladstone was correct, the ability to which dates from the early eighth century discriminate colors had developed only within (Wallace 1927, 69), the 3,183 lines of the poem the past three thousand years. This indicated include sixteen terms descriptive of color, Darwin had been wrong in supposing that which occur a total of thirty-nine times. Most biological evolution proceeded slowly. Until refer to color value, including such words as about 1924, when interest faded, the topic was dark, bright, and pale, together with black, heatedly discussed. Color blindness was white, and gray. conjectured to afflict many ancient and modern Among early dissenters, Grant Allen authors, as well as enormous sectors of the sensibly argued that art is a more reliable population of the world. indicator than language of ability to Geiger, among many who were captivated, discriminate colors. What were being regarded applied Gladstone's reasoning to the as deficiencies in color vocabulary are not Upanishads, the Eddas, and ancient Chinese unique to ancient and tribal languages. Nor are works. He saw evidence that ancient peoples they indicators of atypical vision (Allen 1879, had an inability to distinguish blue. They also 202-21). Allen pointed to Julius Caesar's report had difficulty with green and perhaps with "that the ancient Welsh stained their bodies yellow (Geiger [ 1871 ] 1880). blue," evidence they were not insensitive to the Dr. Hugo Magnus, an ophthalmologist and appeal of that color. Allen distributed a classical scholar from Breslau, "with an twelve-part questionnaire to "missionaries, Aristotelian obliviousness of the possibility of government officials, and other persons testing his statements by experiment before working amongst the most uncivilised races." publishing them," carried the argument further.' He asked every question he could think of He contended that a defective color vocabulary ("Have they separate names for green and blue? characterizes most non-Europeans. According for green and violet?"). Allen concluded that it to him, various visual inadequacies are the is not disinterest but "love for color which cause, of which an inability to see blue is most distinguishes the real savage." common. Field workers set out to get to the bottom of Christine Ladd-Franklin, whose writings on the matter firsthand. W. H. R. Rivers, a color are more widely known than those of member of the Cambridge anthropological others in the group, theorized that ability to expedition to the Tomes Strait, which lies distinguish colors is correlated with racial between Australia and New Guinea, collected superiority (Ladd-Franklin 1929). Members of data on color naming among Murray Islanders. advanced societies see all colors. Nonurban He concluded they could not distinguish peoples ("primitives") perceive the world in between blue and black (Rivers 1901, 44-58). black and white. Or, in the variation more At the St. Louis Exposition of 1904, the psy-

Prime Minister Gladstone and the Blues 195 chologist R. S. Woodworth administered tests (Titchener 1916, 204-36). Colors look different for color blindness to eleven hundred persons under different illumination. Rivers was of diverse racial backgrounds. They were probably moved by expediency: artificial asked to match dark shades of colored paper illumination may not have been available on with pale tints of the same color. Woodworth, the islands in the Torres Strait in 1901. But who failed to recognize this as a test for incli- Rivers saw no connection between the light nation to sort by hue, regarded it as an exami- provided to his subjects and how well they nation of fineness in color discrimination. He were able to make fine color discriminations. found that "Filipinos, and indeed all races The oversight speaks to the issue of his own examined, were inferior to whites in this test. understanding of color. Negritos did better than many more advanced Natives of the Murray Islands, Rivers races" (Segall, Campbell, and Herskovits 1966, reported, had no name for blue and were un- 44). able to make clear discriminations when Though the assumptions of the researchers shown the color (Rivers 1901, 44-58). Titch- seem ridiculous, and they were often misled ener argued that Rivers misunderstood the by their own racism, fallout from the great similes Murray Islanders used to identify color-blindness controversy lingers. A folklore colors. Golegole, for example, is the name of dictionary of recent vintage reports that, "not the cuttlefish and a color. Rivers translated because they are color-blind, but because they golegole as "black," assuming it refers to the have no use for a finer distinction, some black fluid secreted by the animal. Titchener primitive peoples see no more than three or contended golegole makes more sense as an four colors in the world around them. They allusion to the cuttlefish's variegated color. If have no terms for many of the colors other Titchener's reinterpretations are accepted, the peoples see" (Standard Dictionary of Folklore, Murray Islanders made fewer errors in color Mythology, and Legend, 1:242). For the naming than Rivers believed. And they never authors, see implies having a name for, a gross identified colors as black that were not black. though common misapprehension. If they had grouped blue and black together, Everyday experience reveals that linguistic what would the significance of the classifica- differences are not signs of genetic or racial tion have been? Goethe saw the visual similar- differences in ability to distinguish colors. ity between these two colors, though the Americans of diverse ethnic backgrounds all researchers did not. use color names similarly. They all distinguish The theories of Gladstone and his followers blue from green, and both from black. The are interpretations imposed on what were uniformity would not occur if the ability to assumed to be peculiarities unique to non- discriminate colors could be explained by Indo-European languages. If a people applied reference to ethnicity. a common name to both green and blue, the The views of Gladstone and his followers question asked was whether they were visually were subjected to a slowly gathering tide of insensitive to the difference between these scrutiny. Rivers had given color-vision tests to hues. Because syntax encourages the over- Murray Islanders in an old missionary house sight, speakers of English are forever forget- without artificial illumination. The results ting that any color name refers to a range, not were compared to those obtained from British to a few shades of color. Given that the num- observers tested in a well-lighted laboratory. ber of discernible shades of color (about ten The two sets of findings are not comparable, million) greatly exceeds the number of color as Titchener complained fifteen years later names in any known language, vastly differ

196 Prime Minister Gladstone and the Blues

ent colors will inevitably be called by the same that because we are sane the rest of the world name. Brown is an example in English. must be crazy. We commonly see differences in color Gladstone, though forgotten as a color without bothering to say what they are. An theorist, opened a chapter in the history of taste English-speaking observer may appropriately that winds along circuitous paths to settle again use the name green for both an extremely pale on the issue of taste in color, not just the yellowish green and an extremely dark preferences of natives in New Guinea but bluegreen. Only when asked, or compelled by everyone's conception of what constitutes color cir cumstances, do we go beyond the economies harmony and how colors ought to be used. The of everyday word usage to seek terms to taste of an era is not necessarily that of prime express the greater number of color subtleties ministers and anthropologists. But in this case it that can be seen. Language is poorly designed was. Signac praised Delacroix because "he for succinctly explaining the range of variation incited painters to dare everything and not to in continua, including the color continuum. The fear that a harmony can be too brightly colored" difficulty in finding names for exact shades of (Rewald 1943, 16). That, however, was radical color is known and often complained of. talk. From the eighteenth century until well into The research of Gladstone and his followers the twentieth, expressions of a more was flawed in allowing for only two conservative norm are typical. conclusions. The first was that primitive and Goethe thought it "worthy of remark, that ancient peoples were afflicted by inferior color savage nations, uneducated peoples, and vision. The second was that they possessed children have a great predilection for vivid inferior color vocabularies by comparison with colours; that animals are excited to rage by our own. A pathology was invented to explain certain colours; that people of refinement avoid Homer's "black blood." But no comment was vivid colours in their dress and the objects that offered on "black hearted," "yellow coward," are about them, and seem inclined to banish and similar metaphorical forms in English. them altogether from their presence" (Matthaei Wallace felt "tempted to wonder whether 1971, 221). Gladstone was not a bit `colorblind' himself to Santayana reminded us that "children and have appreciated so little the poetic metaphors savages, as we are so often told, delight in embodied in Homer's color epithets" (Wallace bright and variegated colors" (Edman 1936, 1927, 5). The theorists also did not notice that 39). Moholy-Nagy similarly found that dried blood, as a matter of fact, looks black. "children and primitives are particularly Why did variant systems for naming colors attracted to vivid, vital primary colors, attract so much attention? Except among especially red and yellow (Moholy-Nagy [ mathematicians, less interest was taken in the 1921] 1947, 155). Africans, Hawaiians, and unusual methods of counting, often derived Puerto Ricans have each been accused, at from bases other than ten, discovered in many various times and by various accusers, of parts of the non-European world. The answer suspect taste for gaudy combinations of hue, may be that ascription to others of inferior though what is deplored as vulgarism at one conceptions about color is a challenge to their time may become au courant at another. taste. Or it is an assertion of the superiority-- Taste in color is inseparable from taste in not least of all, the moral superiority-of one's art, which is why certain color combinations, own taste. Behind this embarrassing cultural say, pink, orange, and purple, are defended by imperialism lies suspicion of the strangeness pointing to painters (in this case, Matisse) who of the stranger, the unexamined assumption used these combinations. At about the time

Prime Minister Gladstone and the Blues 197

Rivers was testing the Murray Islanders to see (the evidence for sickness is the art), the what could be wrong with their eyes, Oceanic, remaining task is to diagnose the affliction. African, and American Indian art was being Anyone is qualified to do this, and no collected by the European and American examination of the patient is required. ethnographic museums in which much of it still The sickness of an artist who made unloved reposes. Its interest was thought to be just art or the sickness of persons with unusual taste sociological. It was not regarded as art, because in color was not always thought to be it looked different from European art. So did ophthalmological. The issue of moral fitness some European art. Demons were within, not drew attention and continues to be intrusive in just on foreign shores: artists whose work was color theory today. Vanderpoel, observing that deviant but who could not always be written off "the use of agreeable and harmonious colors as charlatans or mentally unbalanced. tends to the sanity of the whole body by The ophthalmologist Ernst Fuchs resisted strengthening the nerves," reported "a few cases explaining art in terms of pathologies of the on record where all sensation of color is eye. Predicting the effect of these abnormali- wanting, everything appearing in differing ties is risky because there are, "for example, " (Vanderpoel 1903, 73). color-blind painters, whose work even with Although hereditary in some instances, vice regard to color stands among the highest in the could not be overlooked because the defect "is art .... It would be a great mistake to impute the also brought on by the excessive use of tobacco, peculiarities in coloration or drawing that are alcohol, and other stimulants." constantly exhibited by certain artists to any Jean-Martin Charcot (1825-93), the French deficiency in their visual sense" (Fuchs [ 1908] specialist in nervous diseases, took an interest. 1924, 244). The argument had to be made Medical science was obliged to contribute because a web of ophthalmological where possible to understanding of the etiology. interpretations had developed. Fuchs was not As Vanderpoel reported, "Dr. Charcot and his impressed by their Victorian pretension to school in have made many examinations verity. But others were and continued to be. into visual disturbances, and through these The human figures in El Greco's paintings, an examinations much of the peculiar coloring and optician confided while fitting my eyeglasses, mannerism of some of the modern painters of are long and thin because the artist had an the so-called impressionist, tachist, mosaist, astigmatism. Van Gogh's Sunflowers, the same gray-in-gray, violet colorist, archaic, vibraist, optician confided, is painted primarily in and color orgaist schools have been explained. yellow because the artist, who had The artists tell the truth when they say that betterdocumented ailments, was afflicted with nature looks to them as they paint it, but they xanthophilia or xanthopsia, because of which are suffering from hysteria or from other everything looked yellow. The condition must nervous derangements by which their sight is have been in remission on those days when van affected" (Vanderpoel 1903, 6). Gogh used other colors in his paintings. Vanderpoel knew enough about color to be What would my optician say if I told him I publishing a book on the subject. Charcot was had no time for an eye exam? I could offer to uninformed about color or art. Despite the send along one of my paintings, for his use in extensive theorizing of the Impressionists, that updating my eyeglass prescription. Even artists are motivated by their ideas about art, persons unfamiliar with the arts need not feel not by nervous derangements, was not a pos- intimidated by an aesthetic theory that is so sibility he considered. Why did Vanderpoel simple. Given the premise that an artist is sick overlook Charcot's limitations? In one of two

198 Prime Minister Gladstone and the Blues

possible answers, his prejudices meshed with or pathological. Whether among the islanders of her own, a meeting of minds that makes strange the Torres Strait or in Leonardo da Vinci, taste allies. In the other, his credentials as a and creativity, a human birthright, are physician, a scientist, led her to assume he reinterpreted as a festering cesspool of could offer superior insights even in areas far abnormalities of eye, mind, or soul. from his field of expertise. Freud rarely used color names in his Charcot's opinion, much quoted, was a writings. Gladstone, when writing on topics watershed in the explanation of taste in color, other than Homer's color vision, probably used color in art, and art. In the early twentieth cen- fewer than are found in Beowulf. The lapse tury, those who looked to pathology for the raises questions about Freud's and Gladstone's key to deviant taste usually cited psychology- eyesight if we ought to apply the reasoning used cal imbalances rather than moral or oph- by Gladstone himself, by Geiger, Magnus, and thalmological defects. Or moral concern was Ladd-Franklin. In the study of human expressed by wondering if the offenders were psychology, others filled in the gap. The Swiss psychologically ill, a method expanded to psychiatrist Hermann Rorschach, rounding out answer any question. Despite his fascination Charcot's suspicion, found that French with the creative process, Freud (who studied Impressionist painting indeed could be equated with Charcot in 1885) became adept at invent- with "colors" (Rorschach 1964, 109). On the ing a fatal flaw to devalue everything from Rorschach inkblot test, undue interest in color artistic genius to altruism to the psychic life of at the expense of form is held typical of women. Impressionist painters and "epileptics, manics, Freud also extended the scope of the imbeciles, paretics, scattered schizophrenics, or method and took more interest in artists notoriously hot-headed, hyper-aggressive and acknowledged to be great than in those who irresponsible `normals' " (Rorschach 1964, 33). attracted a narrow audience. Art became the The message behind this madness is an neurosis and the artist became the patient, autocratic determination to root out summoned to the couch from beyond the grave nonconformity, to set standards. Whether in most cases. The ad hominem aspect of this moral standards alone are the appropriate reckless aggression is smoothed over today. In canon for color and the arts was never certain. the hands of Freud and his followers, the Gustav Fechner and Wilhelm Ostwald, both methodology is called a form of literary trained in the physical sciences, were the most criticism or art criticism. But writers and artists influential of those who hoped to prove that are the targets of investigation, not literature harmony of color and beauty in the arts were and art. The assessments offered become, at governed by scientific laws. The issue of their worst, a high-flying version of the tabloids morality could be laid to rest on the basis that that reveal the secret vices of movie stars and scientific law by its nature is moral, a public figures. Rarely are admirable qualities reasoning that led directly to modern color found in creative workers, or found worth theory. discussing. No responsible commentator argues today The innocuous, though unfruitful, idea that that color in Impressionist painting is best pleasing colors and great art have a moral value explained in terms of nervous derangement. is of earlier vintage than Gladstone. He and his Nor is defective color vision thought to be followers turned the idea upside down. They typical of contemporary artists and heathens set the style of suspecting that what displeases beyond the pale of what Jung called "the white or puzzles any viewer is immoral, depraved, civilized world." But these indiscretions are

Prime Minister Gladstone and the Blues 199 not entirely behind us. Segall, Campbell, and Rivers's tintometer would not have passed Herskovits indicated the great color-blindness muster with Fuchs, who contended "we should controversy "persisted unresolved" in the social not . . . undertake to test the color sense by sciences at least through 1966 (Segall, setting colored objects before the person and Campbell, and Herskovits 1966, 247). It would asking him the name of the color." The problem be comforting to find that lack of reliable is that "an uneducated man will not infrequently information about anomalous color vision call the colors wrong" (Fuchs [ 1908] 1924, contributed to the confusion. But information 247). Titchener, according to R. L. Gregory, was not lacking. It was ignored by the "did a study at Cornell in which respondents researchers, a classic example of self-deception were required to name the colors of bits of in scientific research. paper without the use of abstract color names, Ernst Fuchs, author of a standard text on using instead the names of concrete objects. ophthalmology (1908), and professor of that Analyzing these data in a manner analogous to subject at the University of Vienna, held views Rivers, Titchener obtained specious evidence of on anomalous color vision substantially similar an insensitivity to blue" (Gregory 1966, 46). to those current in professional circles today. Fuchs named Nagel's anomaloscope as the Fuchs found color blindness more difficult to standard clinical instrument. In conjunction diagnose than the anthropological literature of with the Stilling test, it provided "the simplest his day suggests. Rivers, for example, tested and most satisfactory means of determining the Murray Islanders by asking them to sort presence of color blindness" (Fuchs [ 1908] Holmgren wools. But Fuchs found "the wool 1924, 246). Gregory indicated that the test is unsatisfactory," a conclusion with which anomaloscope, although the usual explanation the Optical Society of America more recently of its functioning "must be wrong," remains a concurred. The test "has been criticized as standard test instrument today (Gregory 1966, detecting only about half of the color 125-29). defectives examined, as failing quite a number The anomaloscope was not mentioned by of over-anxious normals, and as permitting Rivers. Nor was the Stilling test, available at the very easy practice improvement" (Fuchs [ time and apparently familiar to the nonmedical 1908] 1924, 245; Optical Society of America public. Grant Allen, before taking issue with the 1953, 140). views on color vision propounded by Another instrument Rivers employed, the Gladstone, Geiger, and Magnus, had tested his Lovibond tintometer, is not a tool for testing own eyes to be sure he was not color-blind for color blindness and went unmentioned by himself. Without prejudice to what Fuchs would Fuchs. Its usual use is in grading colors of have thought about self-testing and industrial materials by comparing light selfdiagnosis, he would have approved of reflected from a sample to light transmitted Allen's report that the test was conducted by through colored glass filters. Industrial users means of "Dr. Stilling's Tables for the criticize the tintometer because the erratic Examination of the Colour-Sense" (Allen 1879, densities of its filters (red, yellow, and blue) 206). make it difficult to calibrate and therefore Judd noted four standard tests for color unreliable (Optical Society of America 1953, blindness used today ( Judd 1979, 204). In the 330). Rivers adapted the machine as a source Holmgren wool test, the subject is asked to of colored patches that the subject was asked match pieces of wool to larger skeins that are to "recogni2e and name correctly" (River red, green, and rose. In the Nagel charts, color 1901, 44-58). spots are placed in a ring on different cards.

200 Prime Minister Gladstone and the Blues

The subject must say which cards are of a single tetaranopia, in which "the blue and yellow hue and which have several hues. On the sensations are absent," the known condition that Stilling charts, spots are arrayed and colored so most closely approximates the blue blindness that they form a digit that can be read by the hypothesized by Gladstone, Magnus, normal observer. In the Ishihara charts, an Ladd-Franklin, and Geiger. Tetaranopia "is improved version of the Stilling charts, some exceedingly rare, since only five cases have digits can be read only by normal observers ever been described" (Optical Society of (trichromats), others only by dichromats. Some America 1953,137). Yet Geiger hypothesized a of the Ishihara cards have two digits, one of universal and total insensitivity to blue-- which will be read by a normal observer, the tetaranopia-among all ancient peoples, while other by a dichromat. Magnus and Ladd-Franklin thought it was The findings produced by Gladstone and his common to most or all non-Europeans. followers were as bizarre as their methods. The I conclude, from reading Fuchs, that Rivers, type of color blindness they most often an anthropologist, and Woodward, a psycho- attributed to entire populations was insensitivity logist, were unqualified to diagnose or assess to blue, sometimes said to extend to include visual anomalies. They should not have been green and dark colors in general. The remote testing anyone's eyes. They knew too little inspiration may have been Goethe, who had about color and color vision and had never conjectured that two people he knew were consulted the ophthalmological literature of unable to see blue. their day. Fuchs's opinions about anomalous Inability to distinguish blue, as a pathology, color vision have stood the test of time better fits uneasily into either current than theirs. ophthalmological theory or that of the early Like Rivers, modern researchers in the twentieth century. Fuchs identified three types social sciences often devise their own tests, of anomalous color vision: protanopia, which hopefully do not incorporate as many deuteranopia, and tritanopia. He defined the misconceptions. Yet misconceptions will first two as insensitivity to, respectively, red and remain difficult to avoid until more people are green. In tritanopia, "probably the primary trained to think visually, to reason in an sensations of yellow and blue are wanting, but orderly manner about what they see. The aes- red and green are present. This condition, which theticizing of color encourages dilettantism, an is hence called blue-yellow-blindness (Hering) assumption that personal feelings are infallible or violet blindness (Helmholtz), is so rare that it and need not be subjected to scrutiny. A thread has been but insufficiently investigated" (Fuchs can be traced, in the history of theories about [ 1908] 1924, 243-44). color, from Gladstone and his followers Gregory agreed that "tritanopia is through Charcot and his followers to Wilhelm extremely rare" (Gregory 1966, 127). The Ostwald and his followers. Research can be Optical Society of America distinguished dead wrong, yet have an effect over a between tritanopia, which is weakened per- prolonged period. If it fails to contribute to a ception of blue and yellow (and is rare), and solution, it becomes part of the problem.

CHAPTER 23 Bassa, Shona, and Ibo

Biologically speaking, there seems little doubt that our colour sense was evolved from our tone sense. Originally the eye perceived only degrees of light. Those who are completely colour-blind show a reversion to this state. Adrian Stokes, Colour and Form

ladstone and his followers might be adjacent at either side. Yet an American written off as ethnocentric, a com- describing it will list the hues as red, orange, mon shortcoming in nineteenth- yellow, green, blue, purple, or something of the century scholarship. The researchers were kind. The continuous gradation of color unsophisticated about color and insensitive to which exists in nature is represented in lan- the arbitrary nature of language. They ques- guage by a series of discrete categories .... tioned societies with different systems for There is nothing inherent either in the spec- naming colors, though no system is more cor- trum or in human perception of it which would rect than another. Yet why should differences compel its division in this way. The specific exist between languages if language does not method of its division is part of the structure of drive perception? We all see the same thing. English" (Gleason 1961, 4). Similarities lie behind linguistic differences. Commenting on color-naming practices in African languages, the modern linguist H. A. Bassa and Warm/Cool Gleason, Jr., reminds us that the spectrum is Bassa and Shona, spoken in Liberia and Zim- "a continuous gradation of color from one end babwe, illustrate other ways of classifying to the other. That is, at any point there is only colors. The Bassa language is unusual in divid- a small difference in the colors immediately ing the spectrum into just two main types of

202 Bassa, Shona, and lbo

color and their subvarieties. The divisions, ziza ward in, say, a watercolor painting, will be and hui, correspond to what are called warm- found seated as firmly on the paper as all other ness and coolness in English. The warm colors color spots if the painting is touched. (ziza) are red, orange, and yellow. The cool The advancing of color from its plane is a colors (hui) are green, blue, and violet. misleading figure of speech. I submit a less Gleason found an advantage to the Bassa problematical explanation. In vision, the eye system for one purpose, simple though the sys- makes sense out of an array of color spots by tem may seem. Western botanists have seeking extremes. We look for the lightest discovered that our method of dividing the spot, the darkest, the brightest, the area of spectrum into six major hues "does not allow greatest contrast. The colors we call warm sufficient generalization for discussion of usually attract the eye more than those we call flower colors. Yellows, oranges, and many reds cool. Poussin often used bright red at the focal are found to constitute one series. Blues, pur- point of the composition of his paintings, a ples, and purplish reds constitute another. These color effective in drawing the viewer's two exhibit fundamental differences that must attention. be treated as basic to any botanical description. Cool colors, given the right circumstances, In order to state the facts succinctly it has been compel attention as effectively as warm colors. necessary to coin two new and more general In Renoir's Madame Charpentier and Her color terms, xanthic and cyanic, for these two Children (Metropolitan Museum), most of the groups. A Bassa-speaking botanist, under no painting is in warm tones of orange and such necessity, would find ziza and hui ade- orange-brown. Two color areas stand out quate for the purpose. They divide the spec- because of their contrast with this scheme. The trum in approximately the way necessary" pale blue of the children's clothing-a cool (Gleason 1961, 5). color-attracts attention for its difference from What do speakers of Bassa notice that the painting's general tonality. The effect is prompts them to group colors in this way? strengthened by the nearby black and white in They attribute significance to the qualities we the dog and the mother's dress and by Renoir's call warmness and coolness, ill-formed and virtuoso brushwork in the children's hair. We confusing concepts in English. Warm colors notice the people and the dog as Renoir (those with longer wavelengths) are said to intended, but particularly the children in their advance visually. Cool colors (shorter blue dresses. wavelengths) are said to recede visually. Chil- In most cases, warm colors, especially red, dren and earnest art students often puzzle, to no attract attention more quickly than cool colors. conclusive end, over whether, say, a red color The key, again, is relationship to a background. spot "comes forward" more than a blue spot In the natural world, the most common colors does. in the environment are the tertiary colors, Red, orange, and yellow, the warm colors, tones of brown and gray. The cool hues of the indeed have a greater tendency to attract the short-wavelength end of the spectrum--blue, eye, especially when set off by dark or dull green, and violet-look visually similar to the colors. But we should not confuse subjective ternaries and tend to get lost among them. For states of mind with action by external agen- this reason, color schemes limited to cool cies: Because the visual field is a two- colors and tertiary colors-say, a room deco- dimensional matrix, colors cannot advance or rated in blue, green, brown, and gray-run a recede. No movement in a third dimension can risk of being considered dull. The colors are occur. The red spot that seems to come for- recognized as similar, perhaps too similar.

Bassa, Shona, and lbo 203

Warm colors provide a higher degree of cipswuka, citema, and cicena, its main generic contrast with the general background tonality. terms for hue or hue/value, Shona, like Bassa, Bright red usually attracts the eye more contains names for more refined variations, quickly than bright yellow, though yellow and subvarieties within the class. These correspond black are a high-contrast pair in traffic signs to English words such as crimson, scarlet, or and the yellow stands out. Whether or not seen rose. against black, we probably equate yellow with The Shona name cipswuka probably would white, because yellow and white are both light. not puzzle speakers of English. Red and violet Bright red, by default, is the color that looks are at opposite ends of the spectrum. But reds most different from other colors, from the and violets look similar, and a range of red- background tonality of the natural world. The violet colors exists "between" red and violet. popular combination of red, white, and The use of citema for both blue and black, of black-strong hue against strong value cicena for both yellow and white, is more contrast-changes in character if bright yellow is surprising. The practice of applying a single substituted for the red. The similarity between name to both a range of hue and either black or yellow and white becomes apparent. white is found in several languages. A cause of Bassa divides color into ziza and hui without controversy in anthropological research, it attributing warmness or coolness to these largely accounts for erroneous reports of color classes. No analogy is drawn between color blindness to blue or of inability to distinguish and temperature. The colors at the blue from black. Color blindness to black was longwavelength end of the spectrum just look never suggested as an explanation, a sign of the different from those at the short-wavelength narrow channel in which the researchers' end. In English, warm and cool are not thoughts flowed. Reports of confusion between regarded as color names, the exact reason white and yellow were also rare, apparently botanists coined xanthic and cyanic. Yet the because the researchers saw the visual similarity words warm and cool function in the same between white and yellow more easily than that manner as, say, red or blue. Warm colors can between black and blue. be sorted from cool colors as easily as red from A single name for blue and black is blue. precluded in English because colors are The English language, unlike Bassa, is struc- segregated by hue, and hue is separated from tured to downplay the importance of warmness value. Blue is classified as a hue. Black, white, or coolness, though the qualities are seen and and gray are identified as values. Speakers of acknowledged. If asked the color of a red Shona treat hue and value as equal in object, the observer is thought to have given a importance, or prefer value. This leads to a wrong answer if warm is supplied. Warm and system with its own refinements. The average cool, like light and dark, are not color names in speaker of English, shown spectral blue and English other than in a de facto sense. A spectral yellow, identifies the pair as blue and speaker of Bassa would find this strange. yellow, which fails to take account of the darker value of blue. Because "the strongest yellow Shona pigment is by nature much lighter, or higher in In the Shona language, cipswuka identifies reds value, than the strongest blue pigment" and purples, grouped together on the basis of (Munsell 1969, 25), giving correct answers in visual similarity. Citema is used for black, blue, English compels the speaker to identify one and blue-green. Cicena identifies white, yel- aspect of the blue/yellow relationship while low, yellow-green, and green. In addition to ignoring another.

204 Bassa, Shona, and lbo

To argue that speakers of English, who Although blue ribbons are awarded as prizes, rarely point out the greater darkness of spec- blue is also equated with negative feelings and tral blue, are unable to see lightness and dark- conditions similar to those associated with ness would be absurd. No more reliable basis black. Blue laws are puritanical (the Puritans exists for the once popular supposition that dressed in black) and designed to curb those speakers of languages such as Shona, with a who do not respect the Sabbath. single generic name for blue and black, see no hue difference between these dark colors. Human experience is richer than human lan- Ibo guage, and when language is used to report In Ibo, a language spoken in Nigeria, color is experience, more is left out than can be initially classified as either o ji or o-cha. O ji included. Language compels speakers to identifies colors that are blackish or dark. Ocha extrapolate, which is both its strength and its identifies colors that are whitish or light. 0-ft weakness. A speaker of English who never and o-cha are not equivalent to the Shona used the word blue without mentioning that citema and cicena, because no separate class blue is darker than yellow would be regarded exists for red-violets. Nor can citema and as eccentric. People would have no idea why cicena be adequately translated as dark and the information was being offered. light, because different conventions govern Segregating hue, the custom in the West, their use. is consistent with Newton's theories about The speaker of English, asked to sort light color. But Newton had a conception of color colors from dark ones, assumes a command to before developing his theories and may have ignore hue. Light blues are placed in one pile, tailored the theories to this conception. In dark blues in another. In Ibo, o ji and o-cha Third World cultures studied before the refer to the value range of the spectral color, advent of television, a tendency to group blue taken as a norm for that color in certain cases. with black is neither unusual nor surprising. Blue is always o ji, a dark color. Yellow is Blue and black look similar, especially if the always o-cha, a light color. Purple and green blue is dull or dark. In The Tibetan Book of the can belong to either class. A dark purple or Dead, five major colors are associated with green is o ji, light varieties are o-cha (figure deities "enhaloed in radiance of rainbow light" 23-1). in a mandala (Evans-Wentz [ 1927 ] 1966,127). The initial division is supplemented by a The colors are white, red, yellow, green, and simile citing an object familiar to Ibo speakers. blue. Black is absent from the rainbow man- An orange object is described by remarking that dala, and blue is the only color of the rainbow its color is o-cha and like palm oil (which is lights with a negative connotation. The soul orange). Red is compared to blood, black to is warned not to be "attracted towards the dull charcoal, green to leaves, blue to the sky, and blue light of the brute-world . . . wherein so forth. Melemele, an alternate name for blue stupidity predominates" (Evans-Wentz [1927] objects, is an exception to the preference for 1966, 130). simile, and identifies blue objects without In the United States, the art of "singing the comparing them to anything else. blues" emerged in a population of African Despite the significance attributed to hue in descent. Feeling blue, in the original form of English, our ability to describe colors would be the idiom, was attributed to the blue devils. severely limited if words referring to lightness Other phrases that metaphorically equate blue and darkness were lacking. The English with sadness may have an independent origin. language includes phraseology similar in con-

Basra, Shona, and Ibo 205

Figure 23-1. Hue (major color) classifications in Ibo. tones, shades, warm colors, or cool colors. Unlike English, which segregates by either hue or Pastels are a range of pale (or whitish) colors, a value but not both at the same time, the Ibo language uses a combined hue/value system. continuum as easily surveyed as the range or continuum of colors called red. Yet pastel, tint, tone, and shade are not regarded as names of colors. Neither are dark colors, grayed colors, pale colors, or other similar constructions. I will be understood if I ask a salesperson to show me dresses in pastel colors. If asked the color of a pink dress, I cannot correctly answer "pastel," which is not the name of a color. A word, say, red, qualifies as a color name in the English language if it identifies a range of colors classified by chromatic difference without regard to lightness or darkness. But the

use of hue as a sorting criterion is not the whole tent to the Shona and lbo terms though story. A color name can also be a word that syntactically quite different. Pastel resembles identifies any achromatic .range from which hue cicena in identifying the class of light colors, is excluded (black, white, gray, light gray, dark including pale pink, pale blue, pale yellow, and gray). What cannot be a color name is a label pale green. Light colors and dark colors are for any range of colors in which value is the used to group in a syntactically similar manner. sorting criterion but chromatic differences are Tint, tone and shade are more formal. They allowed. Black, gray, and white are color too refer to value irrespective of hue, though names. Tone, tint, and pastel are not. The not to value alone. Tint identifies a color that is imperative is to isolate chromatic differences light expressly because it contains white. from value differences, to avoid mixing one Spectral yellow, though light, is not a tint. with the other. Shade identifies a color that is dark because it The bias probably existed in English, or in contains black. Tone identifies a color earlier languages from which modern English is containing gray. Spectral blue, though dark, is derived, well before the hue circles and gray neither a shade nor a tone. Yellow, purple, scales of modern color theorists presented the green, or any other color becomes a tint if separation of hue and value as a discovery. But mixed with white, a shade if mixed with black, it was not the original method in English. a tone if mixed with gray. Because shade also Words such as pale, light, dark, and bright means variety or kind (as in "the exact shade of appear more often in Beowulf than hue names. blue"), tone is sometimes used for a color The primal color names adopted at some later mixed with either black or gray. point might have been regarded as a better Although English provides a facility for system for naming colors. Evidently, the primal classifying colors according to lightness or names came to be classified as the only bona darkness, whether this constitutes classification fide names for colors. Words like light and dark according to color is messy. As in the case of continued to be used, but were regarded as not warmness and coolness, English includes really color names. words that function as color names though we Ibo offers insights into why so many color say they are not color names. Colors can be names in the English language are compressed sorted according to whether pastels , tints, similes, borrowings from the names of objects.

206 Bassa, Shona, and lbo

In Ibo, a color might be identified as, say, light Goethe thought persons of refinement showed a and like a lemon. English has the apparatus for distaste for bright colors. Liking colors is not similar constructions, and we find the quite mainstream, a preference attributed to apparatus, though not the constructions, in women, artists, and aesthetes. I find these ideas Beowulf. "Light, and like a lemon," easily wrongheaded because they grow from a reduces to lemon colored or just lemon. mingling of things that do not belong together. In English, the separation of value from hue Good and evil are in the human mind, as are shores up the idea that black and white should ideas about human sexual mores. Colors, like be regarded as opposites, a moral agenda link- trees, rocks, and clouds, are neither good nor ing the colors with good and bad. The idea has evil. Although the literary (I do not mean no meaning in visual terms, because the pur- literate) cast of the English language prompts us ported oppositeness-as opposed to just in this direction, burdening colors with differentness--cannot be seen. Filling out the inappropriate symbolic associations becomes a picture by grouping the hues into pairs of form of collective baby talk. opposites as well is of more modern vintage, The baby talk is not always harmless. A familiar from nineteenth-century literature on clear definition of what constitutes a color the hues and their complements. Despite a name in the English language would be helpful century of argumentation about color for, say, evaluation of the visual agnosias, the complementarity -- about which hue is oppo- condition in which patients are said to have site to which-we still interpret moral differ- forgotten color names. Developing a definition ences in terms of black and white, rather than is complicated by the double standard, by the in terms of, say, red and green, or orange and presence of pastel, tone, shade, warm, cool and blue. other words that are not considered color English is syntactically inconsistent in names though each names a range of colors. including words that identify ranges of color yet The words cannot be eliminated because we are not considered real color names. Pastel is a need them. de facto color name because the word identifies A more subtle harm is the coarsening of a range of related colors. Its meaning, like that visual sensibility. The Shona doubtless have a of any color name, can only be taught system of morality, evolved without clouding ostensively. The congenitally blind can learn to the issue of whether cicena and citema are say that pastel colors are pale, just as they learn color names. It would be interesting to know to say that grass is green. They can have no whether their moral system is quite as black conception of what either statement means in and white as ours. visual terms; they cannot sort color samples The differing values of the spectral hues are according to whether pastel or green. acknowledged in the Ostwald and Munsell sys- A moral agenda, where it exists, is based on tems. Both systems make clear that, say, no literary concepts. Hue is separated from value to bright yellow can be as dark as a bright pur- preserve symbolic associations that link white ple. But both systems segregate hue from and black with good and evil. Amorality or value. In a more innovative diagram of color immorality has been attributed to bright hue, relationships developed by Kandinsky, white which lacks the moral associations of black is associated with yellow because both are and white. I am not surprised that the Ameri- light, and black is associated with blue because can Puritans and the Calvinists dressed in black both are dark (figure 23-2). Kandinsky may and white, that Santayana believed only chil- have been led to this arrangement by the idea dren and savages liked bright colors, and that that white is a combination of all spectral hues,

Bassa, Shona, and lbo 207

Although yellow and blue, for instance, are presumed to be less different than the conven- tional complements of yellow and purple, I am not convinced that this is the case. How can a child be taught yellow is "more different" from purple than from blue? If a visual criterion cannot be identified, complementar-ity has no meaning in visual experience. Unlike the Shona, we regard blue and black as categorically dissimilar. Signs that the colors are visually related are taken as oddities when noticed. Using, say, Winsor & Newton oil paints, rose madder and black can be mixed to form a series of purples. In mixtures of black and yellow, black simllarly behaves as if it were blue and produces dull olive greens, though not bright greens. A table of reflectances of Figure 23-2. Kandinsky's hue/value scale. Over a artists' pigments, condensed from a compila- period of time, Kandinsky gradually refined his palette to black, white, red, yellow, and blue, presumably because these colors could be regarded as pure, simple, or spiritual. His diagram of relationships among these colors assumes a hierarchy in which hue and value are combined. while black is the absence of light (a presence of no hue). Completing the syllogism, each spectral hue might fall on a continuum between these extremes. If Kandinsky's stepped boxes are rearranged into a straight line, the result is a combined hue/value scale, an integrated system similar to that of the Shona. In another kind of combined hue/value scale, the colors can be arranged in a circle (figure 23-3). Reading clockwise from center bottom (white to black), the colors grow progressively darker. The pairs on the axis show maximum contrast: of value in the case of black and white, of hue in the case of red and green. In this scale no color lies opposite Figure 23-3. Hue/value scale. In this the color identified as its complement on a alternative arrangement of a color wheel, pairs conventional color wheel. But each lies of high-contrast hues lie opposite each other but are not complements in the traditional opposite another of high contrast to it. The sense. Reading around the circle from white to high contrast pairs are white and green, yellow black, the colors grow progressively darker in and blue, orange and violet, red and black. value.

208 Bassa, Shona, and lbo

Figure 23-4. Reflectances of artists' pigments. From "Technical Studies," 1939, Cambridge, Massachusetts. Fogg Art Museum, Harvard University.

c = colors that cannot be matched by monochromatic spectral light.

tion by the Fogg Art Museum, indicates that ent, even if matched closely in value and seen the reflected by ivory from a distance. Among darker varieties of black is in the blue range (figure 23-4). The color, black and navy blue are more easily paint appears black because so little light is confused from a distance than, say, black and reflected. dark maroon. The reason for confusing them is As the table also indicates, the perceived that they look similar. color of paint is not always consistent with the Goethe wrote of blue's "affinity with black." dominant wavelength reflected. Alizarin He found that if "darkness is seen through a crimson is a bluish or purplish red rather than semi-transparent medium, which is itself an orange-red. Yet its dominant wavelength lies illuminated by a light striking on it, a blue in the orange or red-orange range. The colour appears," an observation bearing on the dominant wavelength of Venetian red, a dull Tyndall blues (Matthaei, 1971, 223). orange red, is in the yellow or yellow-orange, Furthermore, "blue gives an impression of rather than red-orange, range. cold, and thus, again, reminds us of shade." Black can be mistaken for blue more easily Goethe's belief that blue is cold need not than for any other color, a sign that the colors have come solely from its visual affinity with look similar. If an area of pale gray (made by black. In English we speak of feeling blue with mixing black and white) is next to an area of cold. And the evolutionary scale, ascended, value-similar pale blue (made by mixing blue seems to show a bias toward muted variations and white), the colors are difficult to distin- on the warm colors, suggesting positive (good, guish from a distance. The combination would warm) affinities for these colors. Although be a poor choice for, say, traffic signs. The blue eyes in human beings and in Siamese cats effect of merging cannot be replicated using are among the exceptions, blue, violet, and pale yellow and pale gray, or pale yellow and green are rare colors among mammalian forms. pale blue. The colors in these sets look differ- Among the warm hues, bright red, orange, and

Basra, Shona, and Ibo 209 yellow are less often seen among mammals silver, red, brown, black, and violet. At depths than among birds, reptiles, insects, and plants. greater that fifteen hundred feet, "all the fishes Muted forms of these colors are common: are black, deep violet, or brown, but the prawns umbers, russet-browns, tans, yellow-browns, wear amazing hues of red, scarlet and purple." and so forth. (Carson 1958, 43). Half the fishes that live in dark waters are luminescent, and some are blind. In deep-sea fishes the cones of the retina Color in Nature tend to atrophy while the rods, which give Attempts to understand color by observing its low-light vision, increase (Carson 1958, 44). occurrence in nature were largely abandoned Evidently these fish, because they have the after Newton. They were briefly revived by retinal cones that provide color vision, Goethe, who pointed to Robert Boyle as his descended from ancestors who lived at model. In biological forms, color guides a shallower depths. When the fish moved into selective ability to absorb light and determines darker, deeper waters, low-light (rod) vision the wavelengths absorbed. This color became more useful. selectivity is integral to the life process, and Even in clear ocean water, no plants live perhaps it was crucial to the origin of life. below six hundred feet, and most live in the The oldest fossil organisms thus far upper two hundred feet. The green discovered resemble modern blue-green algae pigmentation in plant cells, chlorophyll, found (Myxophyceae), which possess blue pigments in small globular bodies called chloroplasts, is in addition to the green chlorophyll found in colorprogrammed to require rays of the warmer other plants. The ancient algae, if also colors. Although various explanations have bluegreen, reflected blue and green rays while been given for the phenomenon, chlorophyll, absorbing and utilizing those in the red, orange, when dissolved in ether, shows a bright red and yellow range. The sea in which they lived fluorescence even if the exciting light is blue or is also blue-green, more greenish near coasts. yellow. (Govindjee and Govindjee 1974). The The rays the algae needed penetrate the sea, but fluorescence is close to the color of human only to shallow depths, as Rachel Carson blood, in which hemoglobin, which is red, pointed out in The Sea Around Us (Carson plays a role similar to that played by 1958). chlorophyll in plants. Red, orange, and yellow rays are available Why have no ocean plants adapted to life in to a depth of two hundred to three hundred feet. the dark depths? Some land plants, such as Greens fade out somewhat lower, and blue mushrooms and fungi, live in the dark. A prob- penetrates to one thousand feet. Violet rays able reason is the coldness that accompanies may reach another thousand feet. Beyond that, absence of light. Although polar bears, all is dark. The red, yellow, and orange rays penguins, and seals live in the Arctic, plant life used by blue-green algae are available disappears at low temperatures. That polar primarily near the surface, limiting the depth at bears have fur is incidental to the more basic which the algae can live. color programming of life forms. In higher Carson found the colors of animals in the plants and animals this turns on the greenness sea generally related to depth. Fish that live of chlorophyll and the redness of hemoglobin. near the surface are often blue or green. Closer Among plants that live on the land, chlo- to a depth of one thousand feet, many crea- rophylls are found in conjunction with other tures are transparent. Between one thousand pigments including the carotenoids, which are and fifteen hundred feet, common colors are red, orange, or yellow. Chlorophylls absorb

210 Bassa, Shona, and lbo

primarily in the red, orange, blue, and violet exposing the child to light, and blue light works ranges. Carotenoids, which account for the better than other colors. color of carrots, absorb mainly blue Coloration in animals, including the redness wavelengths. The carotenes, a class of carote- of mammalian blood, gives every sign of a noids, are converted to vitamin A in the mam- biological purpose, as does the greenness of malian liver. Chlorophyll and the several other plants. Color also seems to have a function in pigments associated with it absorb, in the absence of light, although we know little aggregate, light from most of the wavelengths about this. Human beings assume that color has of the solar spectrum. Houseplants often do no purpose in environments in which human poorly under artificial light, which lacks many beings would be unable to see it. This idea is of these wavelengths. They improve under ill-considered. Red prawns live deeper in the lights balanced to more nearly approximate the ocean than red light waves penetrate and ought colors of light from the sun. to look black to other fish-if the fish perceive The close chemical similarity between chlo- the prawns as we would. The parchment worm rophyll and hemoglobin, the substance that (Chaetopterus pergamentaceus) is strongly gives mammalian blood its red color, is one of luminescent. Yet it lives buried, just its two the most curious color phenomena in nature. ends slightly protruding, in a tube it builds on When cited as an argument in favor of a the sea bottom. No other creature ever sees its vegetarian diet, the assumption is that chlo- orange light. rophyll can be easily converted to hemoglo- Unless invisible, every object is some color, bin because the molecules are identical except even if just transparent or black. We have no for one atom. A porphyrin ring that forms a grounds for assuming that for creatures living in sort of head for the chlorophyll molecule has darkness black is a better color than, say, red or an atom of magnesium at its center. The cor- orange. Interior parts of the human body are not responding porphyrin ring in hemoglobin con- black, even in dark-skinned peoples. Yet these tains an atom of iron instead of magnesium. parts are not exposed to light. Unlike the chloroplasts of plants, red blood The redness of mammalian blood implies cells are not exposed to direct sunlight. absorption of wavelengths in the blue and Mammalian blood moves through the closed green range, a different way of operating than channels of a circulatory system. But sunlight that catalyzed by the greenness of chlorophyll penetrates mammalian tissue through the skin, in plants. Although lower animals lack red as in the process by which it stimulates calcium blood cells and some algae are red, one ques- metabolism and synthesis of vitamin D. tion is whether plants differentiated from Interior parts of the body can also be animals according to whether they utilized affected by light entering through the eye. Sex- primarily the longer or shorter wavelengths of ual maturing in mammals, known to be stimu- the solar spectrum. Whatever the answer, the lated by light, is delayed in congenitally blind life process is intimately associated with the individuals. Richard J. Wurtman, studying colors of the pigments carried by organisms, other light-driven biological rhythms, found not just by light. That coloration is not biologi- that many were color-driven. Green light is cally casual is also implied by its genetic basis. most effective in shifting the body-temperature Each organism, like each part of that organism, rhythm in rats and in suppressing activity of is limited to some colors and not others. the pineal gland (Wurtman 1975, 75) .Jaundice Daisies are never black; human hair is never in premature human infants can be cured by blue or green. Although some living creatures

Bassa, Shona, and lbo 211 are small and others are transparent, no machines have more parts that can break. organism is invisible, another kind of color Against this background of change, color in life constraint. Where life exists, it has a color forms was evidently always genetic. The condition. ancient blue-green algae were all the same Robert Boyle conjectured that white animals color, programmed by this color to live under would be more prevalent in Arctic climates, as certain conditions. Why were the earliest life protective coloration in snow or because less forms not black, capable of absorbing all light was available to "bring out" their colors. wavelengths? Why is neither chlorophyll nor Rarely do assumptions as simple as the second hemoglobin that color, although dead leaves prove out. And the mechanisms of nature often and dried blood turn brown or black? do not conform to what color theory suggests. We rarely associate purpose with the For example, dark objects are better absorbers inanimate world, because the concept suggests of light than light objects, which suggests that consciousness. Yet I doubt that the apparently dark skin would be better for cold climates. purposeful nature of color is unique to life Alternatively, people in cold climates forms. Crystals have their molecules arranged might have developed fluorescent or in an orderly fashion and are often transparent. luminescent skin, since fluorescent and The transparency, which passes light, luminescent colors store light and release it minimizes the deteriorating effect of sunlight. later. People in hot climates might have Do inanimate substances have certain patterns evolved silver skin. Silver is excellent at they tend to preserve? If so, a law is suggested reflecting light. Or all human beings might similar to the law of inertia or Darwin's concept have developed transparent bodies. In the of the preservation of species. tropics, transparency would pass light through. Transparent plastics deteriorate rapidly, In the Arctic, people could use the largely from the effect of ultraviolet transparency of their bodies to focus light, wavelengths. While doing so, some turn yellow, starting fires whenever they wanted. Nature, in a color effective at reflecting or blocking the designing of human skin, operates by ultraviolet. The lens of the human eye, which another logic. Silver, along with luminescence, filters out ultraviolet, also acquires a yellowish fluorescence, and transparency, never appears cast with age, as if this additional assistance in mammalian forms, though common in fish was needed to protect a retina more fragile than and insects. The amount of melanin, a dark in youth. Coal, buried in the ground, is black, pigment, determines the color of human skin. the best absorber of radiant energy. Is this a Protective coloration in animals has been source of the heat needed to transform coal into studied. But no organized body of scientific diamonds? research as broad as, say, cosmology, assesses We know much more about the effect of, the function of color as we perceive it say, vitamins than about how different colors or throughout the physical world. wavelengths of light affect the human body. Whether the universe is one or many, The damage to the earth's ozone layer has been breaking down or developing, biological evo- an unpleasant reminder that excess ultraviolet lution has led to progressively more compli- causes cataracts and skin cancer. The yellow- cated forms, with more complex requirements brown smog over large American cities shifts about the ways in which light or color or both the color balance of sunlight passing through, are used. To grow complex, however, is to to what public health effects we do not fully grow vulnerable, because complicated know.

212 gassa, Shona, and lbo

The color of electric lights is so different and themselves colored only for that reason. from sunlight that the same color camera film This does not explain the mechanism of cannot be used to record both accurately. selectivity in reflectance, of why these objects Except that grow lights are available for plants, are many colors, or colored by a variety of no effort has been made to correct this. Yet we pigments. If every object reflected every know that blue light, weak in indoor illu- wavelength of light, we would live in a world of mination, cures jaundice in premature infants. monochrome white, a reversal of the nominal The correlation bears on the sallow complexion blackness of the night. Nature may tell us more attributed to those who work indoors. about color theory than color theory can tell Newton left us the idea that the objects of about nature. the world can be regarded as reflectors of light

CHAPTER 24 Tristimulus Theory and Metamerism

These persons, as may be gathered from what has been stated, saw fewer colors than other people: hence arose the confusion of different colours. They called the sky rose-colour, and the rose blue, or vice versa. The question now is: did they see both blue or both rose colour? did they see green [as] orange, or orange [as] green? Johann Wolfgang von Goethe, Farbenlehre

olor blindness, a condition that pro- ture of three properly selected lights of vokes controversy among theorists, is different colors. He concluded that the eye no longer cited to support beliefs needed only three receptors to enable all about ancient and primitive cultures. We do colors to be seen. Maxwell developed a coor- not believe grossly defective color vision dinate system for plotting the three lights that afflicts entire populations. Anomalous color became known as the Maxwell triangle (figure vision remains of interest to colorimetrists for 24-1). Hering, who believed the receptors might its bearing on tristimulus theory, an elabora- each be sensitive to a range, identified these tion of the idea that three primary colors exist. ranges as red to green, yellow to blue, and white to black (light to dark). Modern tables of tristimulus value list the Tristimulus Theory relative proportions of red, green, and blue Tristimulus theory is loosely based on ideas light required to match monochromatic lights originally put forth by Thomas Young in the spectral colors for a statistically deter- (1773-1829), lames Clerk Maxwell (1831-79), mined standard observer (see figure 11-2). But and Ewald Hering (1834-1918). Young, an the varieties of anomalous color vision that English physician, discovered that most spec- occur in human populations cannot be entirely tral colors (not all) could be matched by a mix- correlated with what tristimulus theory

213

214 Tristimulus Theory and Metamerism

acuity for green and red is unclear. If, as in pigments, green is a mixture of yellow and blue, normal acuity in the yellow/blue range is inconsistent with impaired ability to see green. Conjectures about what the color blind seen tend to be more confusing than helpful, and the condition is as poorly described as the visual agnosias. Relying on Young's three receptors, three basic types of color vision are identified. These are trichromatism, , and mono- chromatism. Individuals in these groups require, respectively, three, two, or one color(s) of light to match any color they see. Trichromatism (three lights) is considered nor-

mal. But anomalous trichromats, who are Figure 24-1. The Maxwell triangle, All hue variations can be plotted on the Maxwell triangle otherwise asymptomatic, require different between the extremes of red, violet, and green (the proportions of the lights than those found to corners). White lies at the center. (After Maxwell [ be statistically average. Anomalous trichromats 1890] 1965, 121.) may be no more genuinely anomalous than

individuals who require, say, a different implies. The most common type, popularly amount of sleep than is statistically average. called red-green blindness, is an inherited, But questions would arise about tristimulus sexlinked characteristic. The condition is theory if we could show that anomalous sometimes attributed to 10 percent of all men. trichromats do not, in fact, see colors Judd provided a more conservative estimate of 3 differently. percent, along with one-tenth of 1 percent of all women (Judd 1979, 198). No other type of defective acuity for color shows evidence of a Monochromatism genetic basis. The monochromat can match any color with Atypical acuity for red and green would one light, which varies in brightness but not in appear to be a serious impediment to recog- hue. An individual limited in this manner sees nizing traffic lights. Judd contended this was not fewer colors than average and is assumed to so, because a "traffic signal red is a yellow-ish distinguish solely between darkness and red, seen as yellow by the red-green blind; many lightness (black/white). Monochromatism of the green traffic signals have pur-posely been receives less attention in the literature than made bluish green so that redgreen blind dichromatism. But the condition is easier to observers would see the signal as weak blue and reconcile with modern theory about the so be able to tell it from the red signal" Qudd function of the retinal rods and cones. 1979, 198). For Fucks, however, "bluish green The rods, which provide low-light vision, appears to [the color blind] a colorless grey" differentiate primarily between light and dark, (Fucks [1908] 1924, 243). the reason colors become progressively more Because yellow light is said to be a mixture difficult to distinguish in the evening. If the of red and green light, how a traffic light could cones were to become nonfunctional, vision be "seen as yellow" by those with impaired might be limited to dark / light (black /white)

Tristimulus Theory and Metamerism 215 sensations. The eyes of fish that live in the being. At an even lower evolutionary level, depths of the ocean, where little or no light photosensitivity, the primitive response to light reaches, often have atrophied cones. or its absence, can occur even in organisms Theory is not prepared to explain why without eyes. The phototropism of plants, which monochromatism, once attributed to large causes them to grow toward the light, is a sectors of the world's human population, and photosensitivity of this sort. today to dogs and cats, does not occur in isolation in human beings. Vision limited to differentiation between light and dark appears Not Seeing Light and Dark pathological, rather than an otherwise normal Popular beliefs about color infect theory functioning devoid of hue discrimination. It because they are shared by the theorists. usually occurs in conjunction with photophobia Where shall we place diminished acuity for (oversensitivity to light) and low visual acuity colors regarded as "not colors"? Lessened (Fuchs [ 1908] 1924, 244). ability to distinguish black from white (dark Given these conditions, absence of ability from light) is rarely or never singled out as an to distinguish hue is less an anomalous form of impairment of color vision. Yet the condition color vision than a symptom of serious visual actually exists and is far more common than impairments that include inability to classical monochromatism, which may not exist distinguish forms. The legally blind man or at all. woman who can distinguish only degrees of If we forget that form cannot be isolated light and darkness and who has little or no from color in visual experience, we are liable to perception of form is more frequently assume that color vision functions like the encountered than the classical monochromat: tinting applied to black-and-white photos. the individual, if any, who sees the world as if Transparent glazes of color are imagined to it were a perfectly clear black-and-white overlay black-and-white shapes that could photograph. maintain their integrity without the colors. The The human eye without viable cones may model is implicit in the reasoning of Gladstone be as unable to function in a normal manner as and his followers but marginally appropriate the circulatory system without viable red blood even for low-light vision in the evening. Under cells. Yet what constitutes normal functioning daylight illumination, the superior analogue is is arguable in this case. A progressive the color photograph, devoid of black and white degradation in acuity is a feature of vision shapes in the sense that the layers of dye on the under low light conditions. As evening film are red, yellow, and blue. progresses, we lose the ability to see objects Irrespective of how vision ought to be clearly, not just the ability to see their colors. modeled, diminished acuity within the black/ Rod vision is less acute than cone vision and white range is as legitimately a defect of color may resemble the vision of creatures lower on vision as impaired acuity for hue. The condition the evolutionary scale. Some amphibians is commonplace and the subject of an extensive perceive only movement, the reason for the literature. Popularly called night blindness folk advice to stand perfectly still in the (hemeralopia), it consists of more than presence of a snake. normally diminished visual acuity under Creatures able to see only movement must conditions of low illumination. Apparently perceive a monochromatic world of fleeting caused by reduced functioning of the rods, shadows, a condition we would characterize night blindness is often attributed to vitamin as impaired vision if it occurred in a human A deficiency. But it can be symptomatic of

216 Tristimulus Theory and Metamerism

other conditions, and it occurs with the aging and green. But the disparity is considerable. process (Fuchs [1908] 1924, 249-51). Although Red/green maladaptivity is common. not classified as a form of anomalous color Blue/yellow maladaptivity is rare (the literature vision, night blindness implies a lessened ability is divided only on how rare) and associated to distinguish between black and white, with degenerative diseases of the eye. particularly in low-light environments. It The question of whether blindness to blue is coexists with apparently undiminished acuity, possible for an otherwise normal eye turns on under daylight illumination, for red, yellow, whether any visual pigment can be shown to be blue, green, and other colors. uniquely sensitive to blue. If no such pigment Those afflicted by night blindness, or by its exists, there cannot be a pathology-blue milder form manifested by slow visual blindness-brought about by its absence. The adaptation to the dark, see the darkness. Their question remained controversial as recently as difficulty is limited to distinguishing light, a 1957.' white blindness unaccompanied by black Setting aside weakened and absent acuity for blindness. If the condition is caused by impaired the blue/yellow range, respectively tritanopia functioning of the rods, no impairment should and tetaranopia, dichromatism usually implies occur under daylight illumination. I suspect the anomalies limited to the red/green range. condition occurs both by day and by night. I Red/green dichromats are divided into notice, as I grow older, that my eyes adapt more protanopes and deuteranopes. Both are said to slowly to the dark. But I also need brighter light distinguish only between light/dark (black/ for reading at any time. The aging eye may white) and blue/yellow. But protanopes also grow progressively less sensitive to light, in a show an abnormal brightness distribution. To manner too complex to reconcile with Hering's the statistically average viewer, the brightest three simple sets of receptor pairs. The elderly part of the spectrum lies at about 555 do not lose acuity for some colors while millimicrons (yellow/green). For protanopes, it maintaining acuity for others. If the three occurs at 550 to 590 millimicrons receptor pairs exist, can they all age or degrade (yellow/orange). at exactly the same rate? Judd contends that people with anomalous color vision, although classified as impaired, ought to have superior ability to make fine Dichromatism color discriminations in the yellow/orange Colorimetric theory holds that dichromats, range Oudd 1979, 199). But can we assume who require a mixture of two lights to match that the human eye makes the finest any color they see, distinguish black from discriminations of hue in that sector of the white. They also differentiate either blue/yellow spectrum seen as brightest or lightest? or red/green, but not both of these ranges. Although some correlation exists, reading or Because red and green light are said to mix to matching colors under overly strong light is as form yellow light, the condition is curious as difficult as doing so under dim light. described. Diminished vision for the yellow end of the yellow/blue range is inconsistent with full acuity for red and green. Seeing Colors Differently Nothing in tristimulus theory provides a The names people use for colors reflect more basis for predicting that insensitivity to than just what is seen. A color identified as blue/yellow (or to black/white) would occur viridian by one person can be called dark blue- less often than faulty distinction between red green by another and green by a third. This

Tristimulus Theory and Metamerism 217 gives no ground for concluding that the three bespeaks confusion about language rules for observers are having different visual construction of compound color names. experiences. The presumption is reasonable, Orange-green's relatives are topazy yellow and however, that anyone who sees colors blackish white, the terms that inspired Mun- differently will have difficulty understanding sell and the ISCC to demand more rational sys- how to use color names. tems of color naming. Confabulations of this F. H. G. Pitt asked three deuteranopes and type may come from people who pay insuffi- two protanopes to identify the colors of the cient attention to how other people use words. spectrum, and he compiled a table of the names But I doubt that the confabulations can be dis- that were used (figure 24-2). The table would missed as just incomprehensible. Deuteranope be more informative if we knew whether the number three may have followed some private subjects were shown spectral lights, colored logic in concluding that certain colors looked lights, or colored papers thought to reflect somewhat like orange and somewhat like certain wavelengths. And Pitt's own green. predilections may have contributed to the Pitt's companion chart of confusion colors extent to which colors were incorrectly predicts names that might be used for specific identified as yellow. For example, some hue wavelengths or colors by "confused" protanopic intermediaries are included, notably yellow- and deuteranopic observers (figure 24-3). A orange (golden yellow, 583-600 millimicrons) wavelength of 502 millimicrons, for example, is and yellow-green (greenish yellow, 567-578 recognized by a normal observer to be millimicrons). But others, such as red-orange blue-green in color. Pitt predicts that protanopic and blue-violet, are omitted. Even allowing for observers will call the color Nile green, cream, this imbalance, an unusually large number of brown, or purple-brown, because all of these colors are called yellow. For Judd, "the nam- colors look alike to them. ing of colors by a dichromat gives virtually no Pitt's confusion colors show little or no clues to what he is seeing" (Judd 1979, 200). correlation with names used by the dichro- Deuteranope number three applied orange- mats. One protanope identified 502 mil- green, a term meaningless to most people, to limicrons as the area of transition between blue and yellow. The coinage and royal blue. The second identified colors

Figure 24-2. Color naming by five dichromats (three deuteranopes, two protanopes). (Adapted from Judd 1979, 201, after Pitt 1935, xiv.) Description by Wavelength Remarks by Dichromats

Normal Observer (millimicrons) D1 D2 D3 P1 P2 Red 658-780 yellow orange yellow-green dark color-red? orange Orange 600-658 yellow (warm) yellow orange-green yellow lemon Golden yellow 583-600 yellow yellow orange-green yellow lemon Yellow 578-583 yellow yellow (orangy) orange-green yellow yellow Greenish yellow 567-578 yellow yellow (orangy) yellow yellow Green 524-567 greenish- yellow-white yellow yellow nearly light yellow yellow gone, maybe to white red? Blue-green 502-524 whitish pink yellow yellow gone, blue nearly green Blue 431-502 blue mauve blue green royal blue Violet 390-431 blue blue blue dark color violet

218 Tristimulus Theory and Metamerism

Figure24-3Confusion colors for red-green color blind Decades of wild conjecture about what ness. Conjectures about what the color blind see are "primitives" see suggest that inadvisability of not testable.The color names used by Pitt's dichromats are not those predicted as "confusion colors." speculating about what anyone sees, a lesson (Adapted from ludd 1979. 202, after Pitt 1935.1 applicable to predictions about the vision of deuteranopes. An individual of normal vision-a trichromat-is said to require three colors to match any color he or she sees. This is not exactly a demonstration that three colors are needed to produce color vision. Color can be seen by looking at any of the three lights in isolation and the color black is seen when no light is present. Because more mixtures (therefore more colors) can be made with three lights than with two, we might conclude the trichromat sees a greater variety of colors, perhaps ten million rather than seven million. Even that modest assumption was challenged by Edwin Land, inventor of the Polaroid camera (Land [1959] 1961). Land performed experiments in which two black-and-white positive transparencies were made in a split-beam camera. One was taken through a standard photographic filter that passed wavelengths longer than 585 mil- limicrons (red filter, Wratten 24). The other passed wavelengths shorter than 585 mil- limicrons (green filter, Wratten 58). Because of the filters, yellows, greens, blues, and violets looked unusually dark in the first photo, while reds and oranges looked dark in the second. in the blue-green range by the names yellow Using two projectors, each fitted with a that was nearly green or green. Both pro- polarizing filter that could be rotated, the trans- tanopes named a broad range of colors as parencies were projected on a screen in Regis- yellow or lemon, spanning what the testers tration with one another. The long-wave identified as the continuum from orange transparency was additionally passed through through green. Otherwise, they did not give the red filter used in taking it (Wratten 24). The similar answers. And for protanopes, as for composite image appeared on the screen in full anyone else, calling many colors yellow does color. When the red filter was moved to the not mean that all colors given that name look projector with the short-wave image, the exactly alike. One protanope identified both colors appeared in reverse (red became green, extremes of the spectrum (red, violet) as dark, and so forth). For Land, "The departure from but thought the red might be red. The other what we expect on the basis of colorimetry is correctly identified violet, but referred to red as not a small effect but is complete .... Color orange, suggesting that the color looked in images can not be described in terms of relatively red, but not red enough. wave length and, in so far as the color is

Tristimulus Theory and Metamerism 219 changed by alteration of wave length, the from that of black-and-white photography. change does not follow the rules of colormixing Early black and white photographs, taken with theory" (Land [1959] 1961, 388). slow film, amazed people by the fineness of The gauntlet was thrown down. Judd their detail. They matched the acuity of the responded in "Appraisal of Land's Work on human eye, and all details that could be seen Two-Primary Color Projections," admitting that were captured in the photos. Later, fast films even three-color systems fall short of and high-speed lenses produced pictures that reproducing all colors. For this reason, exceeded the acuity of the human eye. They fourcolor reproduction (red, yellow, blue, stopped the action of galloping horses and black) is preferred in the graphic arts. But, Judd hummingbirds' wings. complained, two-color primary systems are not The camera, people said incredulously, new. And none reproduces as full a range of never lies. Photographs are regarded as good colors as a system with three primary colors: evidence in a courtroom. Yet the color pho- "for example, if the scene shows a man in an tograph always lies. No color film has ever olive-drab uniform (Y 4/4), next to a bed of been capable of refined replication of the purple (RB 4/6), with green grass (GY 5/8) colors of objects. This is why painters often in the foreground, and blue sky (B 9/4) in the complain about photographs of their work. In background, about the best that this scene could theory, the three layers of dye used in color be rendered by all possible mixtures of red light camera film ought to make matching "all with incandescent-lamp light would be to show colors" possible. In practice, this does not a man in a brown uniform (YR 4/4), next to a work. bed of purplish-black (RB 1/1) iris, with If, as Judd said, a two-primary system blue-green (BG 5/8) grass in the foreground, produces such colors as "purple-black iris with and pale green (G 7/4) sky in the background" blue-green grass," this bears on what (Judd 1979, 485). dichromate see. Theirs is a system that lacks a Judd's color codes are Munsell notations. third primary. Inconsistently, Judd and others His point is that a reproduction system using have insisted that "the colors perceived by two primaries results in unacceptable photo- protanopes and deuteranopes are the grays and graphic quality. But this is how many photos all kinds of blues and yellows" (Judd 1979, look, notably those produced by Polaroid and 466). An equally likely possibility is a color Instamatic cameras. We tolerate them as universe slightly skewed from our own, a world passable, if not pleasing, "reproductions" of that looks like a Polaroid print. Fuchs found that the scenes that were photographed. The "many of those afflicted with red-green compromise has led to user-friendly cameras blindness are not aware of their defect, and are that no longer need the hours of exposure time much grieved or affronted if they are charged required by early black-and-white films. Faster with it .... In ordinary life, they rarely make film has resulted, however, in progressive mistakes in naming colors" (Fucks [ 1908] degrading of photographic quality over the 1924, 245). past 150 years. We have grown resigned to accepting bad photographs, blurred video images, and the downright ugly colors of color Metamerism television. The Polaroid camera was To match the color of a paint sample other immensely successful, because it gave instant than a metallic of dayglo color, we need not gratification, not quality photography. use the colors mixed for the sample. The The history of diverges phenomenon, familiar to anyone who mixes 220 Tristimulus Theory and Metamerlsm

paints, is called metamerism in the colorimet- Figure 24-4. Metamerism: observer differentiation between simplex and complex Granville grays. The ric literature. The question about anomalous Granville grays have dissimilar reflectances and color vision is why some observers cannot dis- therefore look alike under some viewing conditions but not under others. (Adapted from Kenneth L. tinguish between colors that look different to Kelly, "Observer Differences in Colour-Mixing other observers. Metamerism brings the abso- Functions Studied by Means of a Pair of Metameric Grays," in Symposium on Visual Problems of Color. lute limits into focus. Some pairs of colors can- [New York: Chemical Publishing Company, 19611, not be distinguished by any observer. These pp. 345-63.) colors appear to match under some lighting conditions but do not match under others.

They are mixed from different constituents Observer Eye and Hair Simplex at Angle of and can be shown to have dissimilar reflec- (sex, age) Color 10% 2%

tances. What makes colors look alike that are not M 55 blue, brown red green alike? The color quality of illumination is a M 45 hazel, brown red green likely candidate. All objects look the same F 56 brown, red green brown color in the dark, although this goes M 27 hazel, dark pink pink unrecognized as metamerism. An object that brown M 44 blue, brown red green ordinarily reflects a certain wavelength of light F 39 blue, dark red green cannot do so if no light of that wavelength is brown M 51 dark brown, slightly green present. But metamerism is a function of black pink perspective and not just illumination, which M 41 blue, blond pink green M 38 blue, red pink green implies a mathematical basis. Given sufficient M 21 hazel, pink green distance, no observer can distinguish red from auburn F 50 hazel, brown pink slightly green. A building made of, say, red and green pink bricks appears to be one color when observed M 39 blue, blond pink pink M 62 blue, brown pink green from a sufficient distance. M 34 blue, brown red green Metameric materials developed for use in M 46 blue, brown pink green M 29 dark brown, pink green colorimetric experiments include a pair of black green textiles (prepared by E. I. Stearns) and M 34 dark brown pink pink black the Granville grays, created by Walter M 47 blue, brown match green Granville in 1949. In an experiment designed F 44 brown, dark pink green brown primarily to determine the effect of viewing M 42 brown, pink match distance (angular subtense) on color matching, brown M 48 brown, pink green Kenneth L. Kelly showed two of the Granville brown grays to thirty-nine viewers (figure 24-4). The F 37 dark brown, lavender green dark brown paint used for one of the panels (the simplex F 37 brown, pink pink gray) was a mixture of black and white. That brown F 19 brown, lavender lavender for the other (the complex gray) was a mixture brown including yellow, green, purple, and white. F 29 blue, brown red match F 25 hazel, light lavender match Because of the dissimilar pigments, the panels brown reflect light differently (figure 24-5). Their M 53 blue, brown match green M 52 brown, match green colors look alike under some lighting brown conditions but not under others. F 42 dark brown, pink slightly dark brown pink Kelly's subjects were asked to describe the M 66 blue, brown green green colors of the panels, first at an angular subtense M 68 blue, brown green green Tristimulus Theory and Metamerism 221

M 21 hazel, blond red green greener in daylight than the simplex, and the M 36 blue, blond red green F 18.5 hazel, brown pink pink reverse will be true" at a M 28 green, blond pink pink similar to that of incandescent light, 2,854°K M 38 brown, light pink pink brown (Kelly 1965, 350). Because the testers believed M 56 blue, brown red green that change in angular subtense was equivalent M 31 dark brown, pink pink black to a change from daylight to artificial illumi- M 77 blue, brown green green nation, the simplex gray was predicted to look more reddish at 10 percent (closer distance); of 10 percent and then at a farther distance, 2 more greenish at 2 percent (farther distance). percent. Introducing an additional variable that Whether the predictions were accurate is not should have been saved for another experiment, clear. The experiment, which could have been the lighting was gradually stepped down. simpler, included various manipulations with Subjects were asked to say when it had been filters. reduced to a point that caused the two grays to Kelly reports that younger observers are look alike. Notations were made of the exact more likely to "describe the simplex gray as voltage. Any two colors look alike if seen from redder than the complex at both the 10% and a great enough distance. And any two colors 2% positions," an effect attributed to increase look alike if the lighting is stepped down of ocular pigmentation with age (Kelly 1965, sufficiently, even though, in the case of some 357). Eye color shows a closer correlation, with colors, illumination must be stepped down to blue-eyed observers more likely to deviate complete darkness. from the predictions. The 1931 standard Predictions based on the 1931 standard observer evidently did not have blue eyes. observer (a statistically determined entity) had Twenty-five percent of the responses by been that " the complex gray will appear women offer more detail than the testers

Figure 24-5. Spectral directional reflectances of simplex and complex Granville grays. (After Kenneth L. Kelly, "Observer Differences in Colour-Mixing Functions," in Symposium on Visual Problems of Color, [New York: Chemical Publishing Company, 1961 ] , pp 345-63.) Wavelength Simplex Gray Complex Gray Wavelength Simplex Gray Complex Gray

380 0.317e 0.305e 580 0.343 0.287 390 0.324e 0.307e 590 0.345 0.306 400 0.331 0.309 600 0.346 0.342 410 0.338 0.310 610 0.347 0.389 420 0.337 0.304 620 0.347 0.425 430 0.336 0.301 630 0.347 0.452 440 0.335 0.305 640 0.347 0.480 450 0.334 0.306 650 0.347 0.504 460 0.334 0.316 660 0.348 0.517 470 0.332 0.373 670 0.347 0.516 480 0.331 0.488 680 0.347 0.500 490 0.330 0.528 690 0.348 0.484 500 0.330 0.497 700 0.348 0.483 510 0.331 0.441 710 0.348 0.502 520 0.331 0.391 720 0.347 0.528 530 0.332 0.357 730 0.346 0.556 540 0.334 0.337 740 0.346 0.589 550 0.336 0.311 750 0.345 0.626 560 0.338 0.289 760 0.344e 0.627e 570 0.342 0.281 770 0.343e 0.627e e = extrapolated 222 Tristimulus Theory and Metamerism

required, though limited in this small sample to more difficult, conditions. Imagine that sub- "lavender," or "slightly pink" rather than just jects are able to identify, from a distance of pink. Some of the female subjects evidently eighteen inches, which of two swatches is took greater care in making discriminations black and which is navy blue, or which of two about color or were trying harder to do what yellow-green swatches is more greenish. If they believed was expected. progressively fewer make correct discrimina- tions from distances of four, fifteen, or twenty- five feet, which would likely occur, a basis is A Model for Color Blindness afforded for standards of acuity in differentiat- During the Second World War, Judd replied ing colors. The grading system, like the test, can with polite exasperation to letters from be modeled after that for anomalies of focus. interested citizens. They suggested that if Normal eyesight is labeled 20-20 vision, and dichromats "see colors differently," they might other conditions are identified as deviations be asked to assist the war effort by looking from that. Thus, 20-400 vision indicates an through camouflage impenetrable to the normal ability to see at twenty feet what those of eye. The inquiries, in all their zaniness, are unimpaired vision see at four hundred feet. understandable. Neither the technical nor the Difficulty in discriminating between red popular literature provides a clear model of how and green is more extreme than difficulty in anomalous color vision ought to be understood. discriminating, at a distance, between two Tests are available to reveal the condition. They similar dark greens. Whether some, or all, need to be supplemented by other tests for dichromate are completely unable to fineness of color vision among the general differentiate between any red and any green population. under any circumstances is by no means clear. Color matching is a commonplace task at The literature often implies this is the case. But which some people are better than others. an individual absolutely unable to see a Human beings vary in the required acuity, difference between red and green would use especially under less than optimal conditions. the same name for both colors. The dichromate We can learn much about how to design tests by tested by Pitt did not respond in this manner, considering those available for nearsightedness and no simple pattern emerges from their (myopia) and other defects in form errors (see figure 24-3). discrimination or focus. What dichromate see, if they see colors Given sufficient distance, nobody is capable differently, is a complex issue. Confusing red of reading the letters printed on an eye chart in and green implies confusion about any color an optometrist's office. The myopic individual in which either is a component. It implies has difficulty at near distances. This some degree of confusion about every hue nearsighted observer has no difficulty range, one of several ambiguities the colori- comprehending the difference between E and metric literature does not address. Among F. But the shapes of the letters blur into their others, the assumption that black and white are surroundings, interfering with form not colors breeds the confusing idea that a discrimination. Understanding the difference color sense is not needed to recognize them. between maroon and brown is similarly no Confusing purple with black is classified as in- guarantee against difficulty in distinguishing ability to recognize purple, not inability to between them at a distance or under poor light. recognize black. Although theory does not pre- A test for acuity in differentiating colors clude the condition, we are not asked to con- might present a set of graded, progressively sider the possibility of monochromats who see Tristimulus Theory and Metamerism 223 only blue and yellow, rather than only black green are conditions as loosely defined as the and white. Nor are we told of dichromats with visual agnosias. Vision, contrary to what is normal acuity for red, green, yellow, and blue, assumed in tristimulus theory, is not a set of but none for achromatic tonalities. We are not discrete cubbyholes in which some can be told why the conditions that actually exist closed off without affecting others. Because match up so poorly with what tristimulus the- any hue name refers to a range, all reds cannot ory predicts. What the investigators have be indistinguishable from all greens by even found gives every sign of being conditioned the most extreme dichromat. Between pale pink by what they were looking for. and dark green, a value difference is evident, Should we assume that an individual lack- and the dichromat is said to have no difficulty ing red/green receptors is effectively as blind distinguishing between light and dark. under red light as the normal individual would A more subtle issue is that any color, be if trying to see by gamma rays? Relying on including red, can be mixed with any other. In Judd's description of how the color blind see many cases, the resulting color would not be traffic lights, the answer is no. Those of anom- identified as red, even if red were one of its alous vision see a different color or range of constituents. This presents a serious colors, a surrogate for any range they are una- impediment to imagining a confusion about ble to perceive, The supposition leads to red, or about the range from red to green, thorny questions. How can receptors sensitive which is not potentially a confusion about any only to blue and yellow do double duty by or every other color range. How will a perceiving red (they would perhaps perceive it dichromat see, say, the color of a can of blue as blue) for an eye otherwise incapable of paint into which a quarter can of red paint has seeing that color? If this occurs, the nominal been blended? In theory, the dichromat should blue/yellow receptors should be reclassified as see the blue as if the red had not been added to de facto blue/yellow/red receptors. it. This amounts to a difficulty in seeing blue. Inability to see red or confusing red with CHAPTER 25 Color and Form in Art

Painting as it is now promises to become more subtle-more like music and less like sculpture-and above all it promises color. Vincent van Gogh, Letter #528 The formal relationships within a work of art and among different works of art constitute an order for, and a metaphor of, the entire universe.

Henry Focillon, The Life of Forms in Art

e learn about images by looking at Long before Freud, ancient peoples are said them. We learn about ourselves by to have superstitiously suspected that an object observing how people react to and its image shared a common destiny. They imagery. In his monograph on Michelangelo's were linked by what Sir James George Frazer Moses, Sigmund Freud offered his insights into identified as "sympathetic magic" and a piece of Carrara marble carved to resemble deconstructed in the many volumes of The a man. Michelangelo's shaping of the stone Golden Bough. We join Frazer in scorning such cannot have endowed it with human charac- quaint notions, but we cannot put them entirely ter or personality. Freud would not have cared behind us. I would not enjoy watching a friend to defend the idea that psychoanalytic tech- destroying a photograph of me. niques could be applied to inanimate objects. Psychologists, no more immune than the The inconsistency cannot be smoothed over rest of us to the disease of confusing images by the conceit that Freud was trying to look with objects, are often disturbed by the famil- "through" the stone surface to the person who iar dinner plate on a table (Kohler 1947, had been represented. Because Moses was not 48-51). The plate persists in looking elliptical, the sitter for this portrait, whose face could although everyone knows plates are round. have peered back is conjectural. We often The mind, when quiet, remembers perspective imagine images to be transparent to our prob- and its explanation through optics or projec- ings; to be surrogates for what they represent. tive geometry. No entity can occupy two

224 Color and Form in Art 225 different spatiotemporal locations at the same of paint on canvas (Denis [ 1890] 1966, 509). time, nor can noncongruent configurations be Thus, a distinction could be made between a the same. An image and what that image looks painting and its imagery. The painting, a like cannot be identified with the object that is physical object, could be weighed on a scale. Its imaged. Yet we often behave as if we imagery (or its picture plane) could not be suspected that this should not be the case. weighed, because two-dimensional matrices The issue of what constitutes an image or have no mass. Nor have they secrets behind representation grows increasingly complex, them to be unveiled. There is no "behind" to the driven by technological advances that force us imagery of the eye, the mirror, or the picture to see what was previously unnoticed. plane of a painting. Nelson Goodman was Photographs can be taken of photographs, correct in raising the question of how an artist producing imitations of imitations of imitations can paint the world the way it "is" (Goodman in the Platonic sense. This kind of layering is 1968, 6-10; Gombrich 1960, 297). The task not necessarily aesthetically inferior. W. H. cannot be accomplished. Images are separate Auden's The Old Masters is more interesting, from the three-dimensional world. They neither rather than less so, because the Breughel capture it nor flatten it out so that it can be painting that inspired the poem was inspired by glued to a surface. a passage in Ovid's Metamorphoses. The history of art might be read as a debate Elsewhere, our understanding of northern on the nature of imagery, a dialectic about color Renaissance art is colored by whether we share and form relationships. I suspect the conception the artists' interest in the fantastic imagery of of image making evolved from the discovery of the biblical book of Revelation. constellations in the sky, from imagining lines As Cezanne recognized, art grows from that made pictures by connecting the dot of one previous art, from what we see today in star to the dot of another. Easy steps lead from museums. Studies of iconography in the visual seeing pictures in the sky to imagining pictures arts, like studies of literary sources, propose to drawn on surfaces. Dots can be made on a inform us about who imitated what. They tell surface to replace the dots of the stars in the us about the long chain of what the imitated in sky, or the dots can be eliminated entirely. Each their turn are discovered imitating. can be remembered only as a point at which a Living among images of images of images, line changes direction. we become confused about how to sort one Early peoples would not have picked out the from the other. Australian aborigines were constellations had the sky been without stars, a long ago regarded as naive for wondering link between images and the celestial objects whether the camera that photographed them that inspired them. The development of would capture their souls. Are we more astrology, probably by the Babylonians, rounds sophisticated in assuming that the camera out the picture, elaborating on the idea that captures the aborigines' facial expressions. I some of the constellations, the starry images, think not. Cameras do not capture anything. are meaningful. They affect human destiny. No Their function is to form images, a process vast conceptual leaps are required to move from that should not be personified. the idea of imagining images in the sky to the A color photograph can be taken of any- modern, though literary, conception that images thing that can be seen, because the visual are connected with objects and have significant world can be coded as an array of spots of stories to tell. color in a two-dimensional matrix.For Maurice Is this the process that actually occurred, Denis, a painting was an arrangement of spots or just a process that might have been possi- 226 Color and Form in Art

ble? We cannot know. But the ancient practice was largely an attribute of forms, a skin that of prophesying through looking at the livers of covered them. Leonardo set down his thoughts animals, criticized by Ovid in Metamorphoses, about the colors of objects and about effects similarly seeks to extract meaningfulness from independent of material objects such as solar the random images of nature. Pictures exist in spectra and aerial perspective. artists' minds before being set down on surfaces. The parallel style of criticism, which can be Where could these pictures have come from, traced back to Plato, was overripe by the time except from what was stored, seen, and of Giorgio Vasari (1511-74). Paintings were conceptually rearranged? evaluated by assessing the degree to which the Paleolithic cave painting, the earliest artist imitated or captured the shapes and colors surviving art set down on surfaces, is usually of natural objects. Like Zeuxis' grapes and explained in terms of hunting magic. If the Giotto's fly, paintings were good if they explanation is correct, the idea that images are "looked real," if the two-dimensional image connected with objects is of earlier vintage than seemed a slice of the three-dimensional world. the conviction that they ought to be. Yearning Here is Vasari on Leonardo's cartoon for The for a lost paradise pervades the Greek tales of Battle of Anghieri: Pygmalion (who carved a statue that came to life) and Zeuxis (who painted grapes so Leonardo began a cartoon representing realistically that birds wanted to eat them). the story of Nicolo Piccinino, captain of Nobody living today has seen these fabulous the Duke Filippo of Milan, in which he works of art. Nor can we locate the portrait drew a group of cavalry fighting for a with the painted nose on which Giotto's painted standard, representing vividly the rage fly looked so lifelike that Cimabue tried to and fury both of the men and the horses, brush it away. two of which, with their forefeet The message of these ancient and medieval entangled, are making war no less stories is that an image can aspire to the fiercely with their teeth than those who condition of an object. Somewhere in this ride them. We cannot describe the variety world is the looking glass into which Alice can of the soldiers' garments, with their crests step. The magic looking glass differs from all and other ornaments, and the masterly others in that the world behind it is spatially power he showed in the forms of the similar to the world in front of it and is equally horses, whose muscular strength and accessible. beauty of grace he knew better than any Lewis Carroll may have been just telling other man (Vasari [ 1550] 1957, 154). little girls what he thought they wanted to hear. But his inspired stream of consciousness is The critical inability to separate image from related to Leonardo da Vinci's musings. For object, to recognize that each has different Leonardo, in the notebooks, the visual world attributes, is an inability to think abstractly was a window into, or mirror of, the three- about visual phenomena. As a form of art dimensional forms of the natural world. A criticism, the genre descended to bathos in the collection of recipes is set forth for imitating nineteenth century; to near hysteria, where it the appearance of three-dimensional objects. survived, in the twentieth. Leonardo's self-imposed task was to look Stendhal's negative review of Delacroix's "through" the two-dimensionality of imagery to Massacre of Chios (Louvre) prefigured the a three-dimensional matrix presumed to lie modern idea that if we have the criticism, per- behind. Color, insofar as it entered the picture, haps we do not need the art: Color and Form in Art 227

With the best will in the world, I cannot for a child to imagine a toy soldier coming to admire Delacroix and his Massacre of life but not for an adult art critic to conjecture Chios. This work always appears to be a about what might transpire if a soldier made of picture that was originally intended to spots of paint swung his sword. depict a plague, and whose author, after Yarn spinning of this type is not venal. But having read the newspapers, turned it into it fosters confusion. Toy soldiers do not come a Massacre of Chios. I can only see in the to life. Painted soldiers do not swing swords. large animated corpse which occupies the As Freud failed to remember, stone men nei- middle of the composition an unfortunate ther move nor have human souls. A stone may victim of the plague who has attempted to well have a soul in some animistic sense of its lance his own boils. This is what the own. But that is not the kind of soul Freud was blood appearing on the left side of this seeking. figure indicates. Another fragment that The idea that art copies nature may just no young painters ever omit from their mean that it ought to. That it really does is a pictures of the plague is an infant who dubious conclusion from, so to speak, the tries to suckle at the breast of his already available database. Of, say, the thousands of dead mother; this can be found in the paintings in museums of the crucifixion of right-hand corner of Mr. Delacroix' Christ, not one was executed by an artist who painting . . . . There should have been a had been at the scene. Many premodern works fanatical Turk, as handsome as Girodet's of Western art are literary in inspiration. They Turk, slaughtering Greek women of illustrate stories from books, most often the angelic beauty and menacing an old man, Bible, Homer's Iliad and Odyssey, Ovid's their father, who will in turn, succumb to Metamorphoses, Virgil's Aeneid, and Dante's his blows.' Commedia. If two-dimensional matrices are regarded as Neither Delacroix nor Stendhal had windows into a third dimension, the expecta- attended a massacre. Yet we are asked to con- tion follows that color in paintings ought to sider whether the critic's expectations about look real, appropriate to the objects repre- what massacres really look like are preferable sented. Some liberty was allowed as a sort of to those of the artist. What could have led visual poetic license. That it was limited can be Stendhal, "after having read the newspapers," seen in early objections to the arbitrary use of to the gaffe of imagining that his own imagin- color in French Impressionist and German ings could stake out a superior claim to truth? Expressionist painting. The artists exceeded The critic, who should have known better, was poetic license, straining the unspoken rule that demanding a right to supervise the artist, pictorial imagery ought to approximately though this is not the whole story. adhere to the real colors of objects. Paintings In Stendhal's world, as in Leonardo's or imitate nature, nature imitates the ideal. The Vasari's, images are conventionally read as universe is pictured as if a hall of mirrors, not a surrogates for what they represent. Any critic collection of separable, self-sufficient items without firsthand experience of what was and events. represented was compelled to use imagination A review by Pierre Wolff suggests that the to bridge the gap. No other way was available "unnatural" use of color, rather than the now- to discharge the obligation to enter into the famous broken color, was the disturbing ele- game of make-believe. The discouraging result ment for unfriendly early critics of French is a world in which it is considered infantile Impressionist painting: "Just try to persuade 228 Color and Form in Art

M. Pissarro that trees are not purple, or the sky afterwards. They paid attention to form and not the color of butter; that the kind of things he to color. As early as 1917 I made the students paints can not be seen in any country; and that use the chessboard division for most exercises no real intelligence could be guilty of such in order to free the study of color effects from excesses .... Try to explain to M. Renoir that associations of form" (Itten 1963, 41). Itten's woman's torso is not a mass of rotting flesh, conception of what constitutes form was limited with violet-toned green spots all over it, to configurations invented by his students or indicating a corpse in the last stages of executed with a degree of deliberation. The decay." squares of the chessboard do not count. Pissarro never painted purple trees. Wolff Similarly, the critic Clive Bell's praise of was unable to read the purple as shadow. Or he "significant form" implies that some varieties was led by rancor to pretend he had missed the are not significant. Used in this valueladen point. His objection, in any case, is both to the sense, form becomes an accolade rather than a colors used in rendering objects and to the descriptive label. The usage justifies asking, presumed excess of painting what "can not be elsewhere, whether any form exists in Walt seen in any country." At a much later date, Earle Whitman's Leaves of Grass or in Abstract Loran showed, in Cezanne's Composition, that Expressionist painting. when Cezanne's paintings are compared to For those who make the distinction, what photographs of the scenes, they look different. justifies a sorting of significant form from form By that time, nobody was surprised. that is less significant? Alas, it is usually liter- I cannot see the history of western art, prior ary content. For Henri Focillon, form had to to the twentieth century, as a quest for a copy be more than form, more than configuration of nature. Rather, a nature that had never or how things look. It had to be adorned with existed was invented. For Leonardo da Vinci, ancillary content, meaning, or purpose. How the cracks in a plaster wall could stimulate a things look was not considered interesting. painter to imagine pictorial scenes and images, What counted was the significance the critic a form of free association that precedes the found behind the surface. A firm foundation Rorschach inkblot test by several centuries. The was laid for the conception that we do not American painter Ralph Albert Blakelock need art if we have the ideas of critics about (1847-1919), reasoning similarly, is said to have art. based the composition of Brook by Moonlight Roger Fry was Clive Bell's contemporary. on the reticulated patterning of paint cracks in Yet Fry's ideas were often more visual, more an old zinc bathtub. Neither Leonardo nor modern. Fry's remarks on Cezanne's Houses by Blakelock argued that aggregates of lines could the Marne are art criticism of a different nature be interesting in their own right or need not from the remarks of, say, Vasari or Stendhal: point to a three-dimensional world lying "Behind, a tree divides the composition in half beyond. with the rigid vertical of its trunk, above which At issue is a hierarchical conception of form its foliage forms an almost symmetrical that subordinates it to content or meaning. As pyramid, which is completed and amplified by in the writings of , who taught the group of houses behind. Only to the right is a color class at the Bauhaus, modern thought there a very unaccentuated suggestion of a was often not modern on this point: "In the diagonal movement" (Fry 1958, 61). study of color, I eliminated all searching for Arnheim's thoughts on El Greco's The Vir- form .... The students usually started by draw- gin with Santa Ines and Santa Tecla were simi- ing the outlines of spots and coloring them lar: " The attitudes of the Virgin and the child Color and Form in Art 229

create a slanted axis. The tilt from the upper is praised for his patient conviction that an order right toward the lower left connects the figures can be found in nature. Yet the insight is in the clouds more directly with the saint to the scarcely original. A belief that nature is orderly left" (Arnheim, 1956, 358). The critic looks at is central to Western thought and in all the painting rather than beyond it. A world is probability inspired the builders of Stonehenge. found sufficient unto itself, the coordinate Cezanne's break with traditional rhetoric was system of the picture plane. A painted tree on more radical, as in his observation that natural that picture plane can be the compositional forms all can be seen as constructions based on axis of the painting, although a three- the sphere, cone, and cube (Fry 1958, 52). dimensional tree is rarely or never the axis of Objects are separated from what they mean to us anything. and tied to the cool world of geometry rather Fry's variety of criticism follows the para- than the emotional hothouse of content. The digm of the sciences. His remarks derive their world is spheres, cones, and cylinders, as well validity from their verifiability. If the tree as trees, cows, and people. divides Cezanne's composition in half, then Fry Is the geometry two-dimensional or three- is correct. If it does not, he is incorrect. dimensional? A representational painting is not Stendhal's remarks on Massacre of Chios, usefully regarded as a translation of a three- based on different assumptions, amount to a dimensional world into the two-dimensional pleading that the reader support the critic in world of the picture plane. The visual field is a opinions that are finally unverifiable. For those two-dimensional matrix in its own right. who agree that no picture of a massacre is con- Teaching drawing implies a redirecting of vincing without violated damsels, Stendhal's attention-teaching students to forget about how criticism is valid. For those who disagree, it is they think things really are. Learning to not. What Stendhal sees in the mirror of art is observe how things really look is difficult, a himself, although he may have been passion- battle against acculturation, which teaches a ately convinced it was a mirror of the world. nonvisual method of reasoning. Cezanne, proverbial father of modern art, CHAPTER 26 Subjectivity and the Number of Colors

The fact that the higher vertebrates, and even some insects, distinguish what are to us diversities of colour, by no means proves that their sensations of colour bear any resemblance whatever to ours. Alfred Russel Wallace

very child probably wonders how a beam of light refracted through a many colors exist and decides there is prism), a number of studies have obliged no way of knowing. Estimates by adults scientists to admit that the eye cannot span a wide range. Moses Harris believed there distinguish more than about 180 different were fewer than 660 colors. Birren found hues! (Birren 1969, 50).

there are two different answers .... If Segall, Campbell, and Herskovits, citing Tri- colors and color variations are to be andis, reported that "man can discriminate judged and measured in terms of some 7,500,000 colors" (Segall, Campbell and wavelengths, luminance, degrees of Herskovits 1966, 47; Triandis 1964). And the reflectance, and the like, there are National Bureau of Standards estimated "there millions of colors. But if colors are to be are about ten million surface colors distinguish- judged by the eye and clearly able in daylight by the trained human eye" distinguished by the eye, the variation of (National Bureau of Standards Circular, n.d. a. colors is remarkably limited in 4). The question of how many steps of gray number-and probably to not more than a can be visually identified between black and few thousand. In pure spectral light, for white similarly yields broad estimates, ranging example ( as seen in a spectrometer or in from only 5 or 10 (Sargent), or 9 (Birren), to

230 Subjectivity and the Number of Colors 231

44 (Itten), 214 (Chandler), several hundred his experimental discovery, reported to the (Ostwald), or about 700 (Freeman).' How Royal Society (1671), that "when any one sort would one go about taking a count of one's of Rays bath been well parted from those of the own, in effect as a type of thought experiment? other kinds, it bath afterwards obstinately A tally can be taken of the number of colors, retained its colour, notwithstanding my utmost or of the number of grays, only if that number endeavors to change it" (Whittaker [ 1910) is finite. It is neither finite nor infinite if the 1951, 1:18). Newton construed this as evidence question is metaphysical, like that about the that each individual ray was elemental in the number of angels able to stand on the head of a color it displayed. But this immutability is not pin. A count of the angels cannot be made, not unique to spectral rays. Many or most chemical merely because nobody has ever seen an angel. substances exhibit a color that cannot be The more immediate barrier is that nobody changed. Chlorine gas, for example, is green; knows where to look for one. crystals of copper sulfate are blue. Each retains If the number of colors is infinite, the its color as long as it retains its form, suggesting process of counting them is without end. If no privileged status for the spectral rays in color cannot be located, the counting process behaving similarly. Newton's data merely can have no beginning. Entities are countable suggests that colored light rays are colored. only if the domain in which they exist can be More is needed to argue convincingly that light located. can be equated with color. There are other objections to locating color in light (or to equating one with the other). Locating Color in Light Consider, for example, the proverbial Colors differ from angels in that people explanation that color is seen because rays of regularly see them, or imagine seeing them. light enter the eye after being reflected from The major uncertainly concerns where the objects. The explanation assumes corporeal sighting occurs, an uncertainty that has inspired objects and eyes: it thus suggests that even if an extensive literature. Color may be located in color is light, it cannot be light in isolation. light, the world, the eye, the brain. It may exist Another objection is that the objects light in some, all, or none of these places. It may encounters before entering the eye are not even be located, as I have suggested elsewhere, neutral reflectors. Each displays a selectivity in in a two-dimensional universe unique unto reflectance that determines its color. The itself. selectivity in reflectance is an attribute of the Etymology suggests that color was object, rather than of the light. It can even be originally regarded as a skin or covering that changed by modifying the object. A green table hid the essences of objects, therefore, as an burned to gray ashes no longer looks green, attribute located with those objects. Newton even if no change is effected in the quality of relocated it in light. Or he assumed that color light falling on it. is light, a reasoning irreconcilable with the Among other reasons no explanation of color proposition that color is caused by light. In a vision relies on light alone, the electromagnetic universe in which causality (external agency) scale is merely a scale of hue variation. It is assumed, no phenomenon can be its own cannot explain other components in color, and cause, The assertion that color is light, in other the greater number of colors are neither words, cannot consistently be supported by the assigned places on it nor correlated with arguments that color is caused by light. particular wavelengths. For example, explana- Newton's locating of color in light relied on tions for metallic and iridescent colors typi- 232 Subjectivity and the Number of Colors

cally assume that these colors can only be seen Beyond the fact that the visual field does not by observing metallic or iridescent objects look the way it might be expected to look, (rather than metallic or iridescent light waves). questions about its shape elude answers. Fish, Although locating color in light is far from unlike humans, have eyes that look in opposite satisfactory, the behavior of light presents the directions. According to the theory of binocular primary obstacle to locating it in objects. The (two-eyed) vision, this ought to preclude any color of any object varies according to ability to judge distance. But fish do not forever environmental lighting. In effect, no object has go about bumping into things. Some insects a fixed color independent of the light shining on have multifaceted eyes, in effect, dozens or it. hundreds of eyes. All of these creatures must Many of these difficulties fall away if I am have visual fields subject to different limits only required to count the colors I see, perhaps than those of humans, which cannot, however, because counting itself is a subjective process. tell us what they actually see. Deferring for the moment questions about the The dragonfly's eye, like my own, probably nature of subjectivity. I shall assume that color merges the images from its separate lenses, a is located in the human eye. This is, after all, reminder that what is actually or verifiably in the only theory that allows us to explain why the visual field is not necessarily congruent the blind are unable to see. with what is seen. An excellent example is provided by the default condition for human vision, which mercifully relieves me of the Color Seen by the Eye tedium of continually regarding my nose or the The assumption that color is visual in origin edge of my cheek. Although those anatomical implies that it never exists unless generated by a protuberances extend into the field of vision, I functioning eye, or pair of eyes. Because eyes need not notice them if I would rather not. cannot function without brains, an ancillary Vision, like thought, can be focused by giving neurological apparatus is implied. Through its or withholding attention. association with brain and eye, the visual field The phenomenon of the leading eye simi- is a subjective phenomenon. If the total area larly suggests that consciousness edits, seen by a pair of eyes, it ought to consist, in although the process is not wholly under con- theory, of a pair of circular arenas, each the scious control. Because my leading eye is the field of vision for a single eye. The evidence for left, what I see with both eyes is approximately circularity is that light enters the eye through similar to what I see with only the left eye. I the pupil, an opening that is circular. The lens can verify this by aligning a distant object with of the eye, through which light passes before my extended forefinger and comparing what is reaching the retina, is similarly round. seen with right eye, left eye, or both eyes. I can Receiving rays through these orifices, the no more will myself to make the right eye lead eye evidently either edits or possesses a degree than I can will myself to be ambidextrous. And of insensitivity. The portion of the visual field everyone has a leading eye, just as everyone is seen by both eyes does not appear to be sub- either left-handed or right-handed. stantively different, in either brightness or Among a few certainties about vision, I intensity, from the portion seen by only one eye. know I cannot see everything, sufficient evi- The field of vision for a single eye is not dence that the field of vision is not limitless. perceived as round. Nor is that for both I cannot see what I am not looking at, include- experienced as overlapping discs. ing what is behind me. Nor can vision's range Subjectivity and the Number of Colors 233 be extended beyond what is popularly called the tionally ignored in cataloging colors. The corner of the eye. Straining to look beyond that question of how many different reds exist is corner, I encounter nothing tangible to report. understood to be a question about how many There is only the subjective sense of a machine discernibly different varieties can be found, not refusing to operate beyond predetermined how many individual red spots exist in the limits, unmoved by human effort to compel natural world. more. Number also suggests that an infinite series The eye has a variety of other limits, many cannot be closed, and the spectral hues in this familiar primarily because methods have been sense cannot be infinite in number. In figure found for overcoming them. The telescope and 26-1, numbers on the right side read microscope show what is otherwise too small counterclockwise, increasing from 1 to infinity; to fall within the scope of vision. Eyeglasses the series is 1, 2, 3, 4 . . . n. The circle cannot be correct for imperfect curvature of the lens of closed. There is no final term for n, hence none the eye, improving visual-field acuity. for 1/n either. Seemingly, we cannot count to 1 Stop-action photography and slow-motion beginning with the smallest fraction, just as we cinematography reveal how fast-moving cannot count from i to the end of the number objects (for example, a hummingbird's wing) series, which is essentially linear. would look if the unaided eye could see them. The figure comes to mind in connection The apparently insurmountable barrier is with color value, particularly on such ques- that the visual field can neither be isolated nor tions as whether absolute white or absolute studied directly. Access is solely through black can be reached. It is inapplicable, report of the person who sees, although reports however, to hue. The hue circle can be closed, can be tested in some cases. An optometrist as in the familiar color wheel, and apparently takes note of the error if a patient reading an has no gaps. eye chart incorrectly identifies a letter E as letter D. The same patient's report of discomfort from new eyeglasses is more difficult to assess. If the visual field is limited or finite, it can accommodate only a finite number of colors. This perhaps explains one of the more strik- ing differences between the color continuum and the number continuum. Whether or not because infinitely numerous sets are singular, no integer in the number series replicates any other. Although I have five fingers on each hand, this merely implies I have two sets of fingers, not that there are two number fives. Although Wilhelm Ostwald argued to the contrary,color is not singular in the same sense Figure 26-1. An infinite series cannot be closed. as number. Any given color can occur more Traveling up the right side from the 6:00 position, numbers increase in the series I, 2, 3, 4 . . .n from 1 than once, as in the case of a particular shade to infinity. They decrease from one to infinity along of vermilion seen at more than one time and the left side, where the series is I/1, 1/2, 1/3, 1/4 . . . place. These multiple occurrences are conven- 1/n. In neither case is infinity reached. 234 Subjectivity and the Number of Colors

Differentiating Colors to one observer but different to another, the To count colors, we need to know more than question arises as to whether they should be merely where to locate them and that their counted as two colors or one. Would it be number is finite. Criteria are needed for dis- proper, for example, to count infrared as a tinguishing between one color and another. visually perceptible color if it were discovered They can be easily extrapolated from the crude that a single individual was able to see it? practical methods used in color matching. Standards already exist that suggest an A primitive test might be whether the same answer to the question. The individual who name seems equally appropriate in each case. If excels or sets records is the marker for assess- one object can be called light yellowish green, ing human capability. The mile, we say, can be but another must be qualified as a slightly run in less than four minutes. This merely darker light yellowish green, the two colors are implies that at least one person has done so, merely related rather than exactly the same. even if only once. The limitation of the test is that two colors Relying on the familiar figure of the record properly called by the same name need not setter, a tally of all perceptible colors might necessarily match. Because dark bluish red, for reasonably include all that anyone was able to example, is the name of a class or range of see, even if certain fine differences could be colors, one dark bluish red need not necessarily distinguished only by a single observer, and by match another. that observer only under certain conditions. Color matching is usually accomplished by The criterion of exclusion (I cannot see placing two colored objects in juxtaposition to color A when and where I see color B) pro- one another. It implies addressing the question vides a standard for what constitutes a percep- of whether their colors can be visualized as tible color. It also speaks to the issue of color homogeneous: in effect, as a continuum. If two sameness. If color A and color B can be seen pieces of green paper are regarded in this at the same time and place, both are the same manner and no subjective sense of discon- color. In practical application, if a certain color tinuity is reported by the individual making the swatch is the color of my shirt and also the judgment, both swatches are identified as the color of my neighbor's shirt, then my neighbor same green, either for that observer or for and I have the same color shirt. observers in general. Beyond these relatively easy questions lies We might imagine combining color match- another, more troubling: that of the relation- ing with a variation of the bubble sort used in ship between your perception and mine. computer programming. An ideal observer, shown two colored objects, could be asked to decide whether their colors were the same. Subjectivity Subsequently, the color of a third object might A popular question about the visual field is be compared to those of the first two. At least whether its sensitivity is similar among all in theory, every object in the world could be human beings or even among all higher ver- looked at if the process were continued for a tebrates. Does everyone see the same thing sufficient length of time. The total number of when looking at blue? colors could be determined by taking a tally Given that no two persons have identical after eliminating any color found to be the noses, ears, genes, pulse rates, or electrocardi- same as some other. ograms, it seems fatuous to maintain any There is no ideal observer, and individuals expectation of identical eyes, identical retinas, vary in visual acuity. If two colors look alike or identical neurological processing of visual Subjectivity and the Number of Colors 235

percepts. Even in a single individual, as any colour now appears to all competent observers optometrist and most wearers of eyeglasses can exactly identical" (Allen 1879, 221). The confirm, both eyes may not function similarly. question is not whether Allen's assertion is true, One is often more nearsighted, farsighted, or but how the alleged determination could have astigmatic than the other. been made. We can do more climb into one Some people claim to see color differently another's brains to investigate private percepts with one eye than with the other. Research than we can investigate private thoughts. confirms the existence of this condition, even The question of whether everyone sees the to the extreme defined as pathological. Judd same thing when looking at color is only reports thirty-seven cases of individuals nominally innocuous. Many find it difficult to colorblind in one eye but not in the other.z consider coolly because it bears directly on the No proof can be devised that blue looks the nature of the human condition. Or it is about same to everyone because no way exists of what visual experience can reveal concerning determining how it looks to anyone. This is that condition. sometimes said to be an impassable barrier to If you and I do not see the same thing when the development of a coherent theory of color. looking at green, we share no common How can we communicate with one another perceptions. Thus, we have no genuinely about blue if your perceptual experience of the common experience about which to color cannot be proved identical to mine? communicate. We are utterly isolate from one One answer is that no body of theoretical or another, beyond possibility of meeting. practical knowledge rests on a foundation of To put the question to the test, imagine two proven experiential similitude. It has never observers. Each is asked to look at an apple on a been considered an impediment to the table in a room. The apple, which has a development of geometry that no proof is wormhole, is partly red, partly green. The table, available that a triangle looks the same to perhaps, is blue. everyone. Nor has dentistry suffered from Three questions can be asked about the absence of proof that everyone's toothache event. Will both observers see the room feels the same. interior in the same way? The apple? The color Alfred Russel Wallace, credited along with of the apple? The answer to each question is Darwin for formulating the theory of evolu- no. Visual experience is contextual and tion, pointed out that we cannot assume singular. animals see colors as humans do. Because the The room as an entirety will look different to visual pigments that enable the eye to see are each viewer. Two observers cannot stand or sit known to differ in chemical composition in the same place at the same time and thus among different animal species, the probabil- cannot see from the same perspective. For a ity is that they do not.3 Wallace's reasoning similar reason, two cameras at separate was expanded by Gladstone, Ladd-Franklin, locations in the room will not produce identical and other color theorists who sought to prove photographs. that primitive peoples see colors differently. An apple, like a room, cannot provide two Although the theories may seem bizarre observers with identical visual experiences. Nor today, it is not always easy to agree with those can the limit be transcended. Imagine, for who originally opposed them. Grant Allen example, that observer X stands first at posi- went so far as to contend that "over the whole tion a, seeing an apple that looks unblemished. known world, among the most civilized and A wormhole is discovered by X when he or she the most savage races alike, the perception of moves to position b. 236 Subjectivity and the Number of Colors

Observer Y, traveling a reverse circuit, experiences than it is for two identical movie occupies b, then a, and is led to a different set cameras, filming from different locations, to of inferences. The viewer who initially sees the produce identical film footage. The limitation unblemished side of the apple will be com- is a matter of mechanics before any consider- pelled to revise his or her estimate of its whole- ation of human psychology. ness. The viewer who initially saw the The popular suggestion is that we all see wormhole makes an assessment requiring no the same thing but individually impose differ- later revision. ent interpretations on the common perception. The question of the apple's color follows Actually, we all see different things; there are that of its wholeness. One observer, because of no common perceptions. Interpretation creates his or her vantage point, may conclude that the the comforting illusion that we all see the same apple is mostly green. The other discovers it to thing, are sufficiently similar to understand one be mostly red. Even if the apple's color is another's experiences, or even share a world of uniform, the illumination throughout the room experience (including visual experience) in will not be. Perhaps the side with the wormhole common. is illuminated by strong light. Its color appears Human experience happens only to in- reddish orange, approaching vermilion. The dividuals and is limited to individual location side with no wormhole is in shadow; it looks in times and spaces. I cannot share your alizarin. Two viewers will not see the same experience because I cannot share your flesh or color, any more than they see the same apple. I be at the same place and time where and when perceive a reddish orange apple with a your experiences happen to you. I cannot be wormhole; for your, it is reddish purple with no you. Nor can I see the color of an object hole. exactly as you do. The red apple can be a topic If the visual field is assumed to resemble a of conversation, not because you and I have color movie passing before the eyes of the each seen the same thing, but because we each observer, each individual sees a unique movie know the conventions for using the words red shared with nobody else. Two viewers looking and apple. at the same apple might as well be living in As this incidentally suggests, language has different worlds. For all practical purpose, they an inherent ambiguity that cannot be elimi- actually are. Each perceives a different apple, nated. Red, or any other color name, identifies or a different presentation of the same apple. a range, never a single variety or a single spot Each incorporates the perception into a of color. But this is also true of the apple on different context of past visual experiences. the table. In the range of visual experiences Even if it were possible to assume an identical implied by the term, mine can never be exactly neurological apparatus, it is no more possible the same as anyone else's. for two persons to have identical visual CHAPTER 27 Object or Attribute

We say that something looks heavy or dry or cold, although the eye in fact is unable to know these sensations. But, it is important to notice that we never say anything feels red. William M. Ivins, Jr., Art and Geometry

y well-worn American College Dic- of the word to a concept as removed as eti- tionary shows that color has twenty- quette (good form). Color can be used as a eight meanings, while form has label for color in general or the hues, a limited forty-one. God has only six meanings. Good range. The word can identify all the hues (seen has twenty-six; bad, fifteen; goodness, five; on the color wheel) or the hues exclusive of the virtue, eleven; beauty, three. Faith, hope, and reddish violet range (the hues of the solar charity boast, respectively, nine, eight, and six. spectrum). Even among words with many meanings, an The looseness creates a need for vigilance unusually large number of meanings, not on the issue of which of our many beliefs always consistent with one another, are about color hold for all colors, not just for cer- assigned to only two words. tain subgroups. The statement that color can be correlated with wavelength is meaningful in the case of those hues assigned places on The Words Form and Color the electromagnetic scale. The idea disinte- Form can refer to two-dimensional or three- grates into nonsense when applied to color in dimensional forms, prompting us to forget general. they have dissimilar attributes. The possibility In the case of form, an ad hoc method for for confusion is compounded by application resolving the ambiguous dimensionality is to

237 238 Objector Attribute

select another word for two-dimensional forms. created. Coloring does not answer the parallel Shape is a common choice, used in plane question about colors. Forming (shaping) geometry. Or we can just live with the bespeaks the making or modifying of forms ambiguity and with the shortage of terms (shapes). Coloring only implies the application applicable to two-dimensional universes. of preexisting colors to a preexisting receiving Space-time continuum implies time and three- surface. dimensional space. No comparable term is Through this distinction, forms imply a past available for labeling time and two- in which the process of forming occurred. dimensional space, the space-time continuum of Colors do not, compelling us to regard them as the visual field. givens, existing entities without a history We learn to decipher, in most cases, how explaining how they came to exist. A fractional form or color ought to be understood in a nanosecond after the big bang, cosmologists particular sentence. The task can be difficult, conjecture, forms as we presently know them and dictionaries offer only limited help. did not exist. Did color exist? The theorists Dictionaries list formal meanings rather than imply, as does the Bible, that the early universe what words suggest-that tangled web of probably looked black (not green or associations reflecting societal value judgments. "imperceptible"), a curiosity requiring further A painter or poet whose work is called formless comment. Are we to assume that black, as (as in critiques of Walt Whitman's Leaves of ancient as time and space, came into being with Grass or T. S. Eliot's Waste Land) is not being them at the moment of creation? If not, what praised in most cases. The logic underlying later process accounts for formation of this that use of the word is not made clear in primeval color? dictionaries. The primal color names follow syntactical At a further remove from the scope of the rules similar to that for color. Red is the result dictionary, words acquire personal associations of reddening (of a surface); yellow, a result of that impel each of us to do what we must, as if yellowing; black, of blackening; tan, of tanning following private directives understood by (of the skin). White is the result of whitening; nobody else. Kandinsky's rejection of the gray, of graying; brown, of browning. representational form as a subject for painting Compound color names and nonprimal color is surely linked to the passion for colors he names are rarely used in this manner. Graying traced to his infancy (Grohmann 1958, 29). The has a commonplace meaning; light graying and form/color relationship had a personal meaning dark graying do not. Crimsoning is for Kandinsky beyond what it might mean to occasionally used, but not, say, mauving or the rest of us. tauping. Other exceptions occur. Green is Language communicates through syntax as rarely called a result of greening, although the well as content. On the syntactical level, color usage is permissible. The laundry substance follows different rules than form or shape, called bluing is intended to whiten clothes, not though each term can be used as noun or verb. make them blue. For form and shape, the verb names the For syntactical parallelism with color, form, process by which the noun comes into exis- and shape, we can turn to element, compound, tence. Forms (shapes) result from work, thus and mixture, basic terms in chemistry. The are the end result of a process. Color (verb) chemical elements are primal or given. does not name a process by which color Mixtures and compounds are derivatives (noun) is created. Forming is an appropriate created by combining elements. The syntax for answer to the question of how forms are mixture and compound, like that for form and Objector Attribute 239 shape, presumes performance of physical work: question of what to call chairs that are not blue. a process leading to a result. The syntax for Calling a banana a yellow misleadingly implies element, as for color, does not. that yellow is found only in bananas. The The syntactical forms imply that visually French, in an exception, order black coffee as perceived forms and shapes (like chemical un noir (a black). compounds and mixtures) are made of more primal substances; that colors (like chemical Nouns and Adjectives elements) are modules composed solely of Any color name can be used as a noun or an themselves. The hierarchy reverses that of adjective, a form of syntactical doubling. everyday understanding. Historically, color has Model phrases and sentences can be been regarded as a secondary to some primary constructed illustrating the alternate usages. other than itself. Colors are called unreal or Stress may differ in each pair but meaning subjective, relative to a presumed realness or remains approximately similar. "The red of the objectivity in forms. Color has been regarded as apple" conveys the same meaning as "the red a skin on the surface of objects, an attribute of color of the apple." form or of light. Nouns are the names of primaries: of forms, objects, or objectifiable entities. Adjectives label secondaries: attributes, characteristics, or Rules for Naming Forms and Colors qualities that describe forms, objects, or Rules for naming forms are difficult to identify, objectifiable entities. The language conventions other than that in geometry a regular polygon echo, rather than answer, the question of with more than four edges, say, an octagon, is whether colors are objects or attributes. Colors named by reference to the number of its edges. can be regarded as objects within the confines The implication is that the number of regular of the two-dimensional universe of visual polygons is as great as the highest number in experience. If imagined to exist in a real, three- the number series. Among rules for naming dimensional universe reaching beyond the colors, the borrowing of color names, say, rose, visually perceptible, they become less than from object names rests on an equally simple objects. Colors lack the three-dimensionality rationale. Forms that we see are always necessary for objectivity (self-containment) in a associated with color; particular objects or three-dimensional world. forms are associated with particular colors or Color, though regarded as a perception, ranges of color. People began using object might be thought of as a third dimension that names for color names because whenever they joins height and width in the planar world of looked at objects, they saw the colors of those vision, or a fifth dimension in four-dimensional objects. space-time. The concepts of perception and That few objects are named after colors- dimension blur into one another, a sign that reversal of the mechanism is rare-also has its one or both may be poorly explained. Width, primitive visual logic. Naming an object after a height, and depth, the three spatial dimensions, color other than its own color is tantamount to are initially available as tactile perceptions. The naming the object after an attribute of some twentieth-century conception of time as a other object, a senseless endeavor. But naming dimension departs from the traditional idea that an object after its own color-as in calling a blue a dimension is measurable in miles, meters, or chair a blue-sows confusion about the other units of length marked on a ruler. Time is distinction between color names and object what clocks measure. Color is what the eye names. Calling a blue chair a blue raises the measures. 240 Objector Attribute

Color Names as Adjectives animal, which implies the cat is large in one Whether regarded as attributes or as objects, sense but not in another. Nothing can be green colors are significantly different from other in one sense yet not green in another. Nor can a members of their class. Adjectives are the house have ten windows in one sense but not in names of attributes, which are said to belong to another or be built of stone in one sense but not objects but to be other than objects in their own in another. right. If attributes, color ought to show family Color, number, substance, and other resemblances to other items in that class. separable attributes challenge the convention of Consider, however, the following phrases: assuming the subordination of attributes to objects, a convention reflected in the The tall, beautiful green tree. philosophical distinction between primary and The tall, beautiful green person. secondary qualia. Consider a stone house. Speakers of English are not permitted to call it The adjectives tall and beautiful are a house stone, because the order of the words is relativistic, for they refer to value judgments by important. If a house is made of stone, why is an observer. A tall person is not the same this less significant than its having the shape of height as a tall tree or building, and no criterion a house? For better or worse, we prefer shape to for tallness is equally applicable to each. The substance or color, because shapes can be measure of a person is taken relative to other understood in terms of geometry. people, not relative to trees. I conclude that tall means tall by comparison with other items in a class. Color Names as Nouns Beautiful is similarly fluid, as are canons of Nouns can be divided into two groups. Some beauty. Standards that explain what is meant by identify material objects (such as, dog or a beautiful man or woman are not applicable to house). Others point to entities that are abstract, a beautiful automobile. A beautiful tree is that lack physical substantiality (size, ranked by comparison with other trees, not by patriotism). The groups overlap at two points, comparison with horses. because the border between the concrete and Greenness, unlike tallness, implies compar- the abstract is as vague as that between ison as to color, but never among objects in a substance and vacuum. class. The color of a green person can exactly . At the first point of overlap, we can imagine match that of a green tree, automobile, or any material object as a member of a class of horse. Because a green tree is green according similar objects. To accommodate this, any to an independent color standard, a chart (or noun that labels an object can be used or swatch book) of greens can be assembled, but adapted to identify the genus. Dogs are partic- not a chart of tallnesses or beauties. To see ular animals in "my two dogs are barking," but nothing but green is possible. To see nothing a class in "dogs bark." The class containing all but tall or tallness is not possible, for these dogs is as abstract as the concept of virtue. terms acquire meaning only in relation to Although we can make statements about objects: tall people, tall trees, tall buildings. classes, the classes often cannot be reached. The nonrelativistic function of color is What is nominally research on the dog is shared by number, substance, and various usually data collected by studying three, three other classes of attributes that can be separated thousand, or some limited number of dogs. from objects. A large cat can be called a small At a second point of overlap, abstract ideas Object or Attribute 241 can be talked about as if they had attributes of experience can be reduced to a common the kind unique to corporeal objects. Worries denominator locating it in the world of physical can be called heavy, a figure of speech. objects. This common denominator is the Abstract concepts are not isolated from the movement of objects. When taste, smell, world of corporeal objects, and most can be sound, or tactile sensation occurs, an object in explained as inferences from that world. Eter- the physical world will be found to have nity is the longest possible duration; duration changed its position. Criteria can be established is an inference from changes in objects. These for what an outside observer sees and can include changes human beings observe in measure if a study is made of a person themselves, which they attribute to the aging experiencing one of these perceptions. The process. observer would not share the perception. But Color lacks three-dimensionality, which sharing is not required for verification. I need encourages our regarding it as an ephemera. not fall off a ladder with a friend to confirm The temptation is to group color names with that he or she fell off a ladder. nouns that identify abstract ideas. The classifi- Tactility, taste, and smell, the three most cation is untenable if the defining characteris- primitive senses, are the easiest to assess. Tac- tic of an abstraction is that it can be explained tile sensation occurs when an object in the as an inference from objects. The class of red physical world contacts the perceiver's body, colors can be explained as an inference, drawn often the skin of the hands. The touched object from individual red colors rather than from moves or is moved into proximity with the individual red objects. Any single shade of red perceiver, or the perceiver relocates to the eludes summarization in this manner. What vicinity of the object. If I stand ten feet from a would one have to look at to infer a particular table and neither I nor the table move, any shade of red? outside observer can verify that I have not To the extent that the answer is a particular touched the table-no tactile perception has red object, we perceive redness rather than occurred. inferring it. Inferences are interpretations built Forces touch us and are often reducible to on previous perceptions. Because colors lack a flow of particles, as in the movement of air mass, they impress us as categorically dissimi- molecules called the wind. If an amoeba lar to dogs, chairs, and other physical objects. buffeted by Brownian motion is able to feel, Among shared characteristics, blue, like chair, it feels the particles striking it. We look names an entity we perceive and do not infer. through a microscope, see motion, and infer the As sense perception, color is subjective or force activating the particles. Gravity fits intrapersonal, a special category. We might uneasily with this pattern, because moving par- argue that the referends of color names are nei- ticles have not been identified. I doubt that ther concrete nor abstract. Perceptions are not human beings have a direct tactile apprehen- material objects, not concrete. Neither are they sion of gravity in an ordinary sense of the abstract, because a perception is not an infer- word. I determine whether I am standing in ence. We understand the meanings of concrete water or floating (a form of weightlessness) by and abstract by reference to an external world, whether my feet touch the bottom. A feeling of and color is not located in that world. Perhaps, pressure in my ears as an airplane descends as has been suggested, colors are located in might be attributed to gravity, again not people's heads. directly. I do not recall ever feeling gravity or If we follow this path, every sensory its absence, as one might feel the wind or its experience other than color or visual cessation. 242 Objector Attribute

Taste and smell, specialized forms of touch, Touch, taste, smell, and hearing are share its affiliation with movement. A tasted understandable in terms of simple mechanics, object must contact the perceiver's tongue, a for their source can be traced to the motion of meeting that can be externally verified. In the objects. No explanation of these classes of case of smell, small particles that can be sensory experience is complete without weighed and measured (they have mass) must assuming a physical world in which motion is contact the mucous membrane of the nose. possible and in which objects move. Whether The kinesthetic sense, a superior form of the an individual perception consists of tactility, tactile sense, is sometimes equated with taste, odor, or sound depends on what object empathy or with extrasensory perception. moves and what part of the perceiver's body is Kinesthesia enables me to feel "in my bones" contacted. In the variety of experience I call how to perform the steps of a dance or respond kinesthetic, events occur under the skin. The to the rhythm of events. Testing an individual stomach churns, muscles tense or relax. The experiencing a kinesthetic sensation would subjective impression is of movement within reveal, I am sure, changes in muscle tone, the body, more definitively assessed by blood pressure, brain waves, and other instruments than by an external observer. neurophysiological indicators. Although the brain is sent information by all Explaining hearing in terms of motion is of the sense organs, machines record some more complex. Sound is regarded as a forms of sense data more effectively than movement of sound waves that cannot be seen. others. We have no machines for recording What is seen (and can be verified) is movement tastes, smells, and tactile perceptions. The tasks in the object that is the source of the sound. A are unimaginable, impossible. But recorded hammer strikes a table, a tuning fork vibrates, sounds, say, bird songs, can be an automobile engine begins operating. indistinguishable from the original sounds of Scientific explanations of sound imply that an the natural world. Sound waves are recorded, object in motion establishes a second type of and quite successfully. No waves are motion in a second object. A hammer striking a associated with taste, smell and tactility. table sets up sound waves in air. The sound What we see can be captured, though not as cannot exist without the waves. But the sound faithfully as what we hear. A legend about the waves require the hammer blow. Greek painter Zeuxis says that birds thought Hearing indirectly implies two objects in the grapes he painted looked exactly like real motion, one affecting the other. The motion of grapes. Human beings have more intelligence the first object, the source of the sound, can be than birds. Human beings know the difference seen and provides a basis for prediction. I can between grapes and a painting of grapes, predict you are not hearing a car engine if I see between birds and a photograph of birds. that the engine is not running. If a hammer Holograms record more information than strikes a table near you, I can predict that any photographs, yet not enough. sound you perceived was of a hammer hitting a Human vision-the perception of color- table. Many objects, from engines to piano differs from other senses in its scope. Touch, wires to vocal cords, are capable of making taste, and smell require that the perceived sounds when they move, yet they create none entity touch the part of the human body that when at rest. In a world without motion, no perceives it. I cannot touch or taste what is in sound would be heard. In such a world, would the next room. Sounds in daily life similarly it be possible to see colors? originate in nearby objects. Unless explosives Objector Attribute 243 are detonated or bombs dropped, hearing a before sending it, would ease the burden on the sound five miles away is unusual. Clouds and brain. stars, which we see, lie at much greater dis- A third possibility is that the brain stores tances. Vision is the sense most delicately visual and tactile memories in the same area, in adapted to working across large spatial effect jumbling them together. This could expanses. Indeed, the conception we form of explain why our conception of empty space empty space is visual, though we learn about takes the form of visual images and why visual the emptiness of spaces through tactile images emerge when we consciously try to experience. retrieve memories of tactile experience. Imagining what an empty room looks like is Whenever I try to recall how velvet feels, I also easier than imagining how pacing off the dis- remember what velvet looks like. A common tance around the room would feel. For envi- box for items that ought to be separated may sioning larger empty spaces, perspectival seem a bad idea, a mistake. Among possible effects come to mind. We might imagine, say, purposes, the meld of visual and tactile mem- a landscape with a river flowing toward a dis- ories may feed the kinesthetic sense, a blending tant mountain, an image familiar from the of vision and tactility in which thought flows in backgrounds of Renaissance paintings. To the opposite direction. Watching dancers learn imagine vast spaces, we think of the black of their steps, I feel how I plan to perform those the night sky, which never ends though it may steps. I do not see a mental picture of how I be finite. Or we remember the black within will look while doing so. which we see with closed eyes, another figure Technologically sophisticated experiments for empty space. have provided much information about the Why does our conception of empty space neurological functioning of the brain, an unex- take the form of visual images if three- plored territory. We still know little about the dimensional space is tactile? One answer is relationship of thought to vision and other that memory is poor for tactile experience. I senses, about human ability to reason. Can a have no clear recollection of how empty space brain kept alive without a body think? feels, and the hand is less efficient than the eye Deprived of sensory experience, what would in gauging distances between objects. Another the brain think about? answer is that the eye may have an ability, Western philosophy assumes a categorical albeit limited, to "think," to process the per- difference between the mind and the senses. ceptions of the hand and provide objective The process of reasoning has traditionally been correlatives. If so, it functions like those com- explained somewhat as follows. The senses, puter peripherals, say, printers, that have small which are never able to do anything right, built-in memories of their own to supplement transmit the misinformation they collect. The the main memory of the computer. This mind or brain, like an autocratic paterfamilias, decentralized arrangement could offer practical sorts what is wrong from what is right. How advantages. Pictorial information is complex, a does the brain know what is right if the infor- probable reason for the thickness of the optic mation coming to it is wrong? Well, it just nerve. On a computer, manipulating and knows. The system sounds irritatingly ineffi- storing graphics, say, a page of pictures, cient. In the technocratic countertheory, the requires far more memory and disc space than brain, like the CPU of a computer, is just a cen- a page of words. A limited ability by the eye tral processing unit manipulating input con- to preprocess, to put information in order veyed from peripherals. If this technocratic 244 Object or Attribute

vision is correct, artificial intelligence is not to listen to the beating of my heart. In a possible and human intelligence is overrated. dark room, I cannot decide not to see the color Either scenario is depressing. In one, human of the darkness. Color, however, is explained in intelligence is a metaphysical assumption. In terms of light, an explanation that associates the other, uniquely human intelligence does not vision with the movement of light. This is exist. More optimistic theories are possible. Is inconsistent with the idea that the eye never the brain built on the same model as the eye, switches off, that vision differs from the other which sees black when unstimulated by light? If senses because it cannot be correlated with so, it too should be selfactivating, capable of motion of any object in the external world. functioning when "nothing" is sent to it. This How are these discrepancies to be resolved? small possibility gives life to our conceptions of I would prefer to eliminate the word light intelligence, memory, conscience, soul, spirit, from the English language and use color human dignity. In its favor, the instead. A single phenomenon is confusingly electroencephalogram of a person in a coma known by two different names, and color is the indicates brain activity, though that of a dead more useful name in this case. Setting that issue person is flat. aside, light unquestionably plays a role in vision. This need not mean that light moves, or that we aptly characterize what it does as a The Movement of Light motion. The movement of light is a misnomer, Any machine, natural or man-made, needs an as is movement through time. Each is a external agency to switch it on and off. The traditional conception, colorful, poetic, con- switch for eye and brain, the self-activating fusing. We do not need either one. The habit of organs, is life and death. We begin seeing and saying that light moves creates confusion about thinking at birth or in the womb. At death, the color (which is unmoving), and obscures the brain no longer sends impulses to the more significant issue of what light (and time) electroencephalograph needle. The eyes, though actually do if move is not the right word. they may remain open, no longer see. The familiar litany of state-of-the-art Eyes and eyesight can be lost during life information about light has been much without causing the death of the individual, a popularized in the past twenty years and can be difference between eye and brain. But the brain, concisely recited. The speed of light is a bilaterally symmetrical like the ocular system, constant. Nothing in the universe can move can be grievously injured and continue to faster. Measurements of the wavelength of light function. In some people, half the brain has are extraordinarily accurate, for Loeb and been removed, usually to control epileptic Adams, "among the most accurate physical seizures. The question of why we see blackness measurements known" (Loeb and Adams 1933, when "nothing" exists to be seen reduces to a 440). It thus "becomes readily apparent that in double line of inquiry. The question asks the wavelength of light we have a standard of whether "empty space" is empty, but also why length which is absolutely invariable and exact" the eye is self-activating. (Loeb and Adams 1933, 445). The length of the The function of eye and brain cannot be standard meter has even been redefined "in explained according to the model applicable terms of what we now believe to be an to sense organs other than the eye. We can- invariable unit, the wavelength of light" not identify objects in the natural world that (Jenkins and White 1957, 255). cause seeing and thinking, that turn switches Unlike the movement of a tuning fork, the on and off. I can decide, by an effort of will, movement of light waves or photons cannot Objector Attribute 245 be seen. The movement of light is an abstrac- strating movement for light is quite difficult. tion, an inference from instrument readings. If Light cannot be measured unless the measur-- light moves in any ordinary or experiential ing instrument affects it. To say that light is sense, its behavior should be similar to other now here means an amount has been appre- entities in motion. A reasonable qualification hended by an eye or a measuring instrument. is that human beings are unable to see this The observed portion, because it enters the eye movement because it is too fast, occurs on too or instrument, is precluded from traveling small a scale, or for some similar reason lies onward to there. Interfering with its travels to beyond the scope of human perception. (or from) there is the only way to prove light is If these conditions cannot be met, the here. movement of light is significantly different Seeing light is not evidence that the light is from the movement of other objects. It may in transit, any more than seeing blue is evi- not be movement as the term is generally dence that blueness is in transit. To confirm understood. Imagine a world in which every- movement, the same object must be sighted in thing else moves but light is immobile. Would two places, impossible in the case of light. The this world look different from our own? How portion that finally arrives there is a residue can we tell whether an object has moved? never halted (observed) at any previous here. If The reasoning by which we infer that one chair is seen in Connecticut on Monday, objects have moved implies measurement. A another in New York on Tuesday, we do not chair that has moved or been moved presently properly infer that a chair moved from stands, say, five feet from where previously Connecticut to New York. located. The mode by which it was transported I am mindful of the nature of waves as they is irrelevant. The chair, like a traveling car, travel through, say, water. If a tidal wave might have been continuously visible breaks on a beach, we need not assume that the throughout its journey. As permissibly, its molecules of water on the wet beach were part behavior might have resembled that of an elec- of the wave at its point of origin. The critical tron jumping between orbits, or the science point is that waves in water can be seen fiction voyagers teleported in "Star Trek": the without halting their motion from here to there. chair magically disappeared from one location This cannot be accomplished with light. The and appeared at another. Transportation from movement of light implies a parallelism with here to there is the sole necessity for motion. the movement of objects. Beyond this, what the Here and there must be separated by a mea- term means is unclear, as are the details of the surable distance. journey of light between its present here and a Because movement is transportation rather past or future there. than transformation, a moving object is assumed to retain its physical integrity. A chair can move without losing a leg. Movement The Movement of Time implies a change of location and that the object Light and time are inseparable in modern the- grows older during the moving interim. This ory in the physical sciences. In past centuries aging, conventionally disregarded, is a the movement (propagation) of light was meaningless conception for a ray of light. If believed to be instantaneous, and time was time slows to a halt at the speed of light, rays thought to continue forever. In modern the- of light do not grow older. ory the speed of light is finite. Time slows If we reasonably expect movement to down as it approaches this velocity, "stop- involve relocation from here to there, demon- ping" when the speed of light is reached. 246 Objector Attribute

Time, along with space, is held to date back no inventions, symbols, fables, games with words. earlier than the big bang (Hawking 1988). They shore up the argument that time moves, Despite these revisions, we continue to assume which makes time and change more that time moves from past through present into understandable, less threatening. The metaphor future. Is the movement of time, like the is born out of a human will to fend off the movement of light, a misnomer, a semantic incomprehensible. The conception of time confusion? I think it is, and that some of the moving casts a powerful spell. People imagine ambiguities in our reasoning about color can be they hear it moving in the ticktock of a clock. traced to oddities in our conceptions of time and Attributing movement to time light. anthropomorphizes it, makes it seem more like Nothing moves in an absolute vacuum, a human being. In art and folklore, time is because nothing is present that might move. further reduced to Father Time, an old man Movement is an event. The event happens to with a white beard. If we could plumb a (or requires the presence of ) forms, objects, or collective unconscious to learn the source of material particles, however materiality or the concepts-time, past, future-I suspect the corporeality might be defined. Attributing underlying issue is death. We explain why movement to time and light is peculiar, though death occurs by saying each person has an the conceptions are traditional. In neither case allotted time and time passes; time runs out for can a movie camera photograph the movement that person. The past must have been invented in slow motion, as might be done with, say, the to explain where the dead go. The future can be movement of a galloping horse. Time involves defined as our fate, our own deaths. The words no particles that might move and has no origin fate, fatal, and fatality all have the same root. or destination. It lacks the verifiable here and The three Fates, in Greek mythology, were not there needed to demonstrate movement in the concerned about the future of the universe. usual sense of the word. Their job was to spin and cut threads that Language facilitates talking about past and represented the life and death of individual future as if they were other times. But we are human beings. We have not experimented with unable to reach or transport ourselves to either. the idea of jettisoning these metaphorical The only time that exists for living human security blankets and confronting the beings is the present, from which we cannot incomprehensible without preconceptions. escape. The admonition to live in the present is Nobody has ever seen time or observed it unnecessary. We have no choice. Time travel is moving. But movement has positive connota- a science fiction fantasy, possibly because no tions in human terms. It separates the quick time exists other than our own. from the dead, suggests progress and mastery, We do not know where the present goes provides important metaphors. People act off- when no longer the present, or where a light icially at formal meetings by making motions. goes when turned off. The most painful conun- The papers filed to commence a lawsuit are drum posed by our conception of time is rarely moving papers. Experiences that affect us are discussed but may lie behind the yearning to also called moving. Life is metaphorically understand the nature of time. Human beings called a journey, a movement rather than a have never understood what happens to the series of changes. We deplore people whose dead, where they go, or whether they or the lives "go nowhere," who have "no direction." past exist in any understandable sense. Human beings want to feel that the human race We have as little idea of where the future is on the move, going somewhere, marching is, because past and future are ancient human along with time, keeping up with the times. Object or Attribute 247

Time has a single here, which is now, and a recall the past to demonstrate that if the year double there, the past and the future. The 1666 were recycled, Newton would perform "arrow of time," a modern coinage, expresses the experiment "again." Historical events can the idea that the movement of time is not be documented, not replicated. Theories about reversible, though subatomic particles, we are the past, including Darwin's theory of evolu- told, can move in a reverse direction through tion, are constrained by the same limit. They time. Early human beings may have found the cannot be proved or disproved according to arrow of time less significant than we do. the standards of proof for laboratory exper- Their myths, collected by Sir James George iments. Frazer in the many volumes of The Golden Ornate theories about the nature of proof Bough, stress the cyclical nature of life, not the have sought to reconcile this discrepancy. The idea that the world moves on. In the cyclical simpler answer is difficult only because it cuts conception, an analogy drawn from the cycles across many disciplines. I suspect we cannot of the natural world, people, animals, and prove things about the past or the future vegetation live and die. New life emerges from because past and future do not exist. Time does the remains of the old, a series of changes in not move, or move from one to the other. A which nothing changes. more apt term is needed to describe what time Human beings search for a purpose in life, does, or to explain what we mean by time. suggesting that any purpose is not easily obvi- Light does not move and probably cannot be ous. In Buddhism, the world of changes is shown to exist as an independent entity regarded as Maya, illusion. All forms, past, separate from color. If the word light were present, and future, are illusion. Effectively, eliminated from the English language, our the only time is this instant, in which we are ideas about it could be rephrased by talking experiencing the illusion. The Judeo-Christian about color. Some adjustment would be neces- tradition and the humanistic tradition, unlike in sary, but no loss in clarity would occur. Clarity, many ways, identify the future as the sig- I think, would increase. We would refine nificant time, the time when the purpose of life thought by eliminating nonvisual concepts, by will be fulfilled. If the future is a convenient bringing our ideas more into harmony with the fiction, a figure of speech, both are lost in a phenomenological world. dream. For the believer, the future is the The conception of time and space as a con- moment when each of us will be judged by our tinuum is a triumph of twentieth-century creator. For the humanist, the future is the time thought. It implies that time and space both when human beings will have fulfilled human move or neither moves. We are forced to potential, completed the self-appointed task of choose either of two models for time. In the mastering the universe. simpler model, assume that the matrix we call Interesting questions arise about time in the present includes all time and all space. connection with scientific proof. If white light Time does not move (nor does light), because passed through a prism is separated into the no past exists that time could come from, no spectral rays, anyone at any time should be future to which it could be going. Clocks can able to repeat Newton's experiment to confirm exist in this world, but an interval must be this. Proof, in the physical sciences, means the defined as a specified number of ticks of a original experiment can be repeated to yield watch. The ticks can be counted, though no the same result. explanation can be given for what happened to How can we prove that Newton performed the ticks already heard. The present, in this his prism experiment in 1666? We cannot conception of time, has features similar to 248 Objector Attribute

those of the black holes imagined by modern (entropy) always increases in systems. The cosmologists. Nothing escapes from the second law of thermodynamics once meant that present, certainly not light. Tomorrow never everything breaks down sooner or later; no arrives. Like the horizon, it keeps receding, and perpetual motion machine can exist. If it is we draw no closer to it. Tomorrow and evidence of the direction that time is movingor yesterday are fables, like the pot of gold under that time is moving-we need to know in which the rainbow. direction the arrow should point. One person's For the alternate model, assume that past idea of order is another's conception of and future are meaningful concepts, rather than disorder. If the early universe was an ocean of convenient figures of speech. We can imagine radiation and quarks, why is this state more time racing from one to the other as if it were orderly than its present state, with intelligent the Orient Express. In that case, space cannot life on earth? consistently be static. If time and space are a The second law of thermodynamics makes continuum-stuck together-space must move sense in its original form. If adapted to tell into the past in stately tandem with time. We where to find the future, it makes less sense. shall have to accustom ourselves to talking My car is not 100 percent efficient; it wastes about the arrow of space (a parallel to the arrow energy. From the perspective of the entire of time) and to thinking of space moving, as universe, however, the energy wasted by my car does time. I find this model too confusing. It is recycled, as will also eventually be the case burdens us with an infinity of space-time with my car. Furthermore, systems break down, continua, one existing in the present and the but not like the wonderful one-horse shay that others at points in the past and future that we deteriorated at the same rate in all of its parts. If can never reach. a car breaks down because, say, its brakes fail, Stephen W. Hawking identifies three the rest of the car can look perfect. If I wanted arrows of time, each pointing in the direction in to monitor the signs that the universe will not which time is moving. These arrows are our endure forever-the signs at which the evidence, from the physical sciences, that time thermodynamic arrow is pointing-where are is moving. these signs, and what are they? We remember the past, not the future, the The third arrow of time, the cosmological psychological arrow. This arrow, which of the arrow, points in the direction the universe is three has the greatest appeal to the imagination, expanding. Whether time will run backward if can be misleading. Sometimes we do know the universe contracts has been asked. Unless what will happen in the future (I know we know the universe will expand forever, the tomorrow will be Tuesday), but this is not cosmological arrow is not a reliable guide to called "remembering the future." We do not where the future is located. We can remember the past entirely, and no person can consistently believe that the universe is remember another person's memories. In the life expanding and also that time does not move, of the individual or the race, more is forgotten that no time exists except the present. than remembered. Some things that are remembered never happened. The thermodynamic arrow identifies the Fermions and Bosons future as a period of greater disorder than the The Copenhagen interpretation of quantum past. Its foundation is the second law of ther- theory allows us to imagine multiple realities modynamics, which holds that disorder and sometimes requires that we do so. These Object or Attribute 249

multiple realities can take the form of, say, different types of receptors are required to swarms of ghost electrons that exist in the relate to the two types of particles. form of probabilities and coalesce into a single electron when observed. Universes may exist parallel to our own, in which what might have Object or Attribute occurred has occurred. Vision differs from other perceptual expe- These are wonderful conjectures. They do riences in that the colors we see cannot be not touch on the multiple realities we live with explained in terms of an immediately verifiable every day: the three-dimensional world of movement of objects. If light moves, we do not touch, the two-dimensional world of vision or see the movement. Relying on current color, and thought, devoid of spatial extension conceptions in the physical sciences, color and although we regard it as existing in time. Can a vision are boson related. Other sense percep- three-dimensional universe include other tions are fermion related. Different classes of universes that are not three-dimensional? If it particles are involved. can, why is it impossible to tell the color of an If we assume that color exists in a world of apple touched in the dark? If touch, vision, and its own-a parallel reality like those of the thought are parallel realities, what glue holds physical sciences-the question of whether color them together? is attribute or object can be answered. Assume Possibly, the visual field is three-dimen- that objects can be isolated, can exist sional, or usefully regarded in that way. We independently. Isolating implies separating an can assume that the third dimension is color, entity from other entities in its class, though a substitute for the depth of the world of not necessarily from those in other classes. touch. The distinction between fermions and Colors are separable in this sense. Blue is dis- bosons, in quantum mechanics, speaks to the tinct from red, which is why I cannot see both issue of why we have no organ that can both at the same time and place. Any individual see and touch. Fermions, named after the color can be isolated from other colors. In this statistics developed by Enrico Fermi and Paul sense, all colors are objects. We may want to Dirac, obey what is known as the Pauli exclu- avoid confusion by calling them objects of per- sion principle. They move incrementally, from ception. one quantum state to another. Protons, elec- Are colors separable from three-dimen- trons, and neutrons are fermions. Bosons (after sional objects? The blueness of a chair cannot the Bose-Einstein statistics), do not obey the be severed from the chair. We cannot transport Pauli exclusion principle and can occupy the the chair to one location and its color to same energy state. Photons, the massless par- another. To explain this limit, we might regard ticles that account for color and light, are the color as if it were glued to the chair. I think bosons. this is an insufficient model. Any chair can be The distinction-roughly, a distinction located in three-dimensional space by means of between matter and energy-suggests that the the x, y, z coordinates familiar from high two-dimensional world of vision (color) is a school geometry. Its color cannot be placed in universe of bosons; the three-dimensional that manner, because vision is limited to two world of touch, a universe of fermions. Human spatial dimensions. I conclude that the color of beings evidently lack a neurological apparatus a chair is not part of the chair in a three- capable of relating to both at once. No single dimensional sense, and that this is why nobody sense both touches and sees, perhaps because can determine by touching a chair whether its 250 Objector Attribute

color is red or blue. The color is in another less real, than three-dimensionality. As a world from the chair, in a different coordinate result, the English language is not rich in terms system. Color lies in the two-dimensional that encourage sensibility to two-dimensional image of the chair that we see rather than in the matrices, to the abstract aspects of visual three-dimensional chair that we touch. experience. And metaphors that refer to flat- Certain traditional questions about color, ness or two-dimensionality almost always have visual imagery, and vision devalue them in negative connotations. subtle ways. We are asked to consider, say, Being without depth or having nothing whether color is real, whether images are below the surface is not admired in poetry, ephemera, whether our eyes deceive us. These people, or ideas. Being without substance is questions have been asked for centuries. I sus- also bad, although being spiritual is good. pect their slant reflects societal attitudes about Novelists are criticized if their characters are two-dimensionality. Plane geometry is not two-dimensional. Voices without expression regarded as inferior to solid geometry, an are flat, as are musical notes sung off key. Yet exception. In most other cases, two- seeing objects is as important as bumping into dimensionality impresses us as less significant, them. CHAPTER 28 Conveying Information about Color

The Munsell System, as adapted to problems of color notation and description, has become the most highly developed 'tool' to be developed anywhere. Thanks to the cooperation of numerous scientists, to the work of scientific committees, to the endeavors of such organizations as the Inter-Society Color Council, the Optical Society of America, the National Bureau of Standards, the System today is pre-eminent. Faber Birren, A Grammar of Color

any color names exist, and new Nomenclatures and Notational Systems coinages are introduced daily. But Familiar examples of scientific nomenclature the primal terms have remained sta- are found in chemistry and the biological ble over an extended time. Few generalizations sciences. Each consists of an alternate set of can be made about the large body of names that names for objects identified differently on an are not primal, other than that most are everyday basis. For the chemist, table salt is borrowed from the names of objects. The sodium chloride. For the ornithologist, the primary rule for learning color names derived Eastern robin is Turdus migratorius from object names is that the viewer can refer migratorius. The technical terms are derived to the object if uncertain about the color. The from Greek and Latin, a mechanism intended to meaning of any color name is best clarified by prevent their being confused with ordinary ostensive definition, by showing a sample of English words. the color or color range. A swatch or sample is Within a nomenclature, each individual mandatory if a color is to be matched. Despite scientific name is a code, conveying informa- claims to the contrary by color theorists and tion about relationships among objects in the colorimetrists, no scientific nomenclature for system. The terms sodium chloride and color exists. Nor does any system approach sodium hydroxide indicate that each of these one. compounds has sodium as a constituent, a

251 252 Conveying Information about Color

family resemblance unrevealed when the language names and not sufficiently different substances are identified as salt and lye. from them. The names range from Munsell's Biological names provide similar clues. Robins blue-green to the National Bureau of and bluebirds are related, because both belong Standards's moderate bluish green and light to the family Muscicapidae. grayish reddish brown. When Munsell developed his notational Although similar names are used in the system for colors, he apparently intended to several systems, they are not applied to colors provide color with a nomenclature similar to in a similar manner. No international agree- that used in chemistry. The same path was ment has been reached on whether one sys- followed by Ostwald, who was trained as a tem is preferable to another. In the United chemist. But the Munsell and Ostwald States, the Inter-Society Color Council and the notational systems, like those of their followers, National Bureau of Standards have evolved a fail to provide a distinctive set of alternate standardized system based on Munsell's (ISCC- names. The color known in ordinary language as NBS). In Great Britain, the comparable system red has no scientific name other than red. When is founded on Ostwald's methods and names. James Clerk Maxwell and other physicists Standard blue is not the same standard, and not describe light of various wavelengths, they the same blue, on both sides of the Atlantic. identify the colors by such everyday terms as When names in a nomenclature system dif- red, green, and scarlet. These terms are of fer from those used every day, this insulates limited usefulness. Adults rarely make errors in them from the loose and changing usage of identifying objects as, say, red or blue. Yet, as ordinary language. Ambiguity, for example, is Munsell complained, confusion reigns when we not unique to color names. Nor is the tendency want to convey the minute color differences to proliferate-. Many plants and animals are between a particular shade of blue and hundreds known by dozens of different popular names. of others that are similar but not exactly the The same name may be used to identify differ- same. ent species by people in different locales. A Because no scientific nomenclature for color single species may thus acquire several names. exists, describing the exact color of a beam of What has been understood in developing the monochromatic spectral light is as difficult as nomenclature of the biological sciences is that describing the exact color of any other colored trying to edit or organize popular names is an object. Arnheim, citing a listing complied by exercise in futility. The names, as a group, have Hiler, notes that the color of light with a no sense or order and usually developed in a wavelength of 600 millimicrons "has been random manner. With sufficient effort, much described by various authors as Orange could be learned about how people use the Chrome, Golden Poppy, Spectrum Orange, terms cattail or bluish green. But this is not the Bitter Sweet Orange, Oriental Red, Saturn Red, expeditious way to systematize information Cadmium Red Orange, and Red Orange" about plants or colors. Assigning the botanical (Arnheim 1956, 348). No scientific way to use name Typha latifolia to the cattail and providing color names exists, because no scientific color a description of what the plant looks like places names exist. it in the nomenclature of botany without The shortcoming of the major cataloging opening that other can of worms: the question systems is less an absolute lack of names (the of whether people always call cattails cattails ISCC gives names for 267 major color classes) and consistently use the name only for that than the character of the names that were plant. chosen. These names are variations on ordinary Colors, admittedly, are a special case. We Conveying Information about Color 253 cannot describe what they look like and thus value and chroma, measured on a scale devised cannot pair descriptions with names. The visual by Munsell and specified for each color in his difference between blue and yellow eludes system. A typical notation might be R 5/10, words, although a botanist can concisely state identified by Munsell as a shade of vermilion the difference between a cattail and an oak tree. (Munsell [ 1905 ] 1961, 20). The hue symbols Colors are said to have fewer than half a dozen are as follows: qualia, a probable reason for the difficulty of the task. Hue, value, chroma, and glossiness are R red most frequently mentioned. Plants have a YR yellow-red greater number of defining attributes. Many of Y yellow these attributes are countable and measurable, GY green-yellow including the number of lobes to a leaf and the G green manner of branching of stems. BG blue-green The defining characters of plant species B blue would not be recognized as such by an PB purple-blue untrained, though interested, observer. Yet they P purple can be pointed out by the technically trained. RP red-purple Whether colors can be described depends on whether they have qualia beyond those usually A chemist on the East Coast can send a cited and whether these qualia are countable or chemist on the West Coast the formula for a measurable. We also need to reassess those chemical substance. In a similar manner, one situations in which measurement is thought to colorimetrist can write to another thousands of have occurred. Whether anything pertinent has miles away. Munsell notations can be used to been measured is questionable if we do not identify a color being discussed. Munsell hoped know whether light with a wavelength of 600 to convince us that the chemist mailing a millimicrons is "Orange Chrome, Golden formula and the individual posting a Munsell Poppy, Spectrum Orange, Bitter Sweet Orange, notation are engaged in similar activities. But Oriental Red, Saturn Red, Cadmium Red the respondent receiving the Munsell notation Orange, [or] Red Orange." will be unable to determine, by reading it, what the intended color looks like. A copy of the Munsell color book is required to locate the Munsell Color Notation swatch corresponding to that notation. Among familiar chemical symbols, Na, Fe, O, This is the fatal flaw in the system, which H, and CI respectively identify atoms of provides ostensive definition, masquerading as sodium, iron, oxygen, hydrogen, and chlorine. an abstract notational system. A chemical In chemistry, these and symbols for the other notation, say, NaCl, tells what a substance is. A elements can be combined into formulas that Munsell notation, say, R 5/10, does not tell display the result of chemical combination. The what a color looks like. It just indicates where chemical notation for water, H20, indicates that to look to find out. The Munsell notation the molecule consists of two atoms of hydrogen resembles a literary citation. It tells where to (H), one of oxygen (O). locate information but does nc I provide the In Munsell's notational system, following information directly. the format used in chemistry, letter symbols The entire tedious notational system that are provided for each of ten major hues. These Munsell recommends can be eliminated by are supplemented by numerical indications of sending a color sample. All concerned can be 254 Conveying Information about Color

mercifully relieved of any need to look up the industrial user, a figure whose shadow swatches in books. Even if the books are looms large in color theory. Theorists from retained, a notational system is not necessary Goethe onward pleaded the relevance of their for identifying individual swatches. Each of the ideas about color to dyers and other practical 1,200 samples in Munsell's Book of Color can persons (Eastlake [1840] 1970, 289). The be assigned a number from 1 to 1,200. A swatch untutored wisdom of the simple soul was labeled number 376 is as clearly identified as asserted against the authority of science. This another marked R 5/10. Swatches can be wisdom was also to stand against the equally useful if labeled, say, "page 5, second obdurateness of artists, who were not row, third swatch from left." unanimous in their gratitude for color systems. In the glossary of A Color Notation, Munsell Goethe's praise of the dyer Jeremias Friedrich contends his system provides "an exact and Giilich is unaccompanied by any clear specific description of a color, using symbols explanation of what is praiseworthy in the ideas and numerals, written HUE/VALUE/ Gulich published (Eastlake [1840] 1970, 291). I CHROMA" (Munsell [1905] 1961, 59). suspect Goethe was no more deeply interested Elsewhere, the claim of precision is qualified. in dye technology than in the chemistry of Color notation is just "a very convenient means ceramic glazes. Goethe had no praise for for recording color combinations, when Newton and praised Giilich instead. pigments are not at hand" (Munsell [1905] The modern industrial user, still courted in 1961, 28). The system would be more literature explaining why the ISCC-NBS convenient if not tied to the availability of the system is needed, is more sophisticated than Munsell Book of Color. the simple dyer of Goethe's day or the honest Munsell argued that his notation provided tailor to whom Chevreul dispensed advice on an "exact and specific description of a color." the use of color in clothing (Chevreul [1845] But an additional requirement must be met to 1980, 165-79). This industrial user is not well develop an abstract notational system. The served by any of the systems. Imagine a description must not be ostensive or must be scenario in which a hypothetical manufacturer direct rather than indirect. Identifying a par- is sent a Munsell notation. This industrial user ticular color as the blue used in the wallpaper of is instructed to ensure that all sweaters made in the master bedroom at the White House in the factory match the color indicated by the Washington, D. C., is admirably exact and notation. Having located, in the handy Munsell precise. This wallpaper, like the Munsell Book Book of Color, the swatch to which the notation of Color, is available for public examination by is keyed, the manufacturer must determine how those who find a reason to take an interest. We to proceed. understand, however, the prudence of not Removing the swatch from the book to pass sending respondents far afield to study swatch it along to the dyers means it will be unavail- books or wallpapers. Those who want a color able if needed again. Sending the entire expen- matched exactly will send a sample. If color sive book means it will be unavailable and may combinations must be recorded and "pigments be mislaid. If I were the manufacturer, I would are not at hand," the pigments are easier to make up color swatches or have them made. come by than the Munsell Book of Color. How is time or effort saved by the notation? The avoidance of Greek and Latin coinages The communicator bears certain burdens in systematizing the naming of colors was in the art of communicating. If I wanted meant to cater to the presumed limitations of another person to understand an unfamiliar Conveying Information about Color 255 word I was using, I would explain what the and related printing processes lies the crudity of word meant, I would not hand the person a color photography. Perhaps because color dictionary. If I wanted the person to understand sensitivity in camera films has been sacrificed exactly what color I had in mind, I would pass for speed, color films in wide usage along a sample of the color. I would not send (Kodachrome, Ektachrome, Kodacolor, and so the Munsell Book of Color or expect a copy to forth) have narrower sensitivity than the human be procured. Munsell turns this societal norm eye to color nuance. Even the best color upside down with no benefit that I can see. photographs fall short of matching the colors in The purported usefulness of Munsell paintings or in color swatches. Silkscreened notations to industry rests on the assumption swatches tipped into books by hand provide a that the chain of command consists of only two better opportunity for color control, therefore people, each of whom has on hand the requisite greater potential for accuracy. Like other book of color swatches. In manufacturing colored materials, these swatches are affected anything, many people are involved, and any of by time and exposure to light. them may need to know the color of the object. Do chemical substances exist that reflect a Supplying each with the Munsell Book of Color single wavelength of light? Could, say, a is uneconomical and unnecessary. If samples of standard yellow be defined as the color of only one color are needed for the job at hand, crystalline sulfur? These directions have not no reason exists for every person involved to been explored, and the standardizing of color have samples of twelve hundred colors. has not been related to the standardizing of, say, the standard meter. The standard meter, a platinum bar kept in Paris, has been redefined in terms of the speed of light, regarded as invariant. A meter is now the distance light Color Standards travels in 0.000000003335640952 seconds. If Even if Munsell's proposals were not light is just another name for color, the standard questionable on theoretical grounds, meter has been redefined in terms of color. We implementing them would not be easy. rarely think of length in these terms. Commercial and technological factors To standardize the colors of the spectrum complicate the task of making color standards in a manner similar to that used for metric available. Modern VGA monitors for computers length, a single wavelength might be chosen are capable of displaying 256,000 colors, a small as, say, standard red. To dispose of Goethe's number compared to the ten million colors that complaint that light is not color, a standard red the National Bureau of Standards computes can color might be defined as the color of this light be visually distinguished. Paper swatches of reflected from a standard white surface, say a standardized colors are expensive to produce and surface coated with barium sulfate or titanium do not reproduce well in ordinary books. The oxide. Difficulties intrude in this process. If a limitations of technology are part wavelength in the red range were arbitrarily of the difficulty. Despite the vigilence of chosen as standard red and that light reflected museums, we are all accustomed to seeing, from a white surface, the illumination in the printed in books, reproductions of famous room would have to be specified as well. Also, paintings in which the color reproduction is red light is defined in physics as a range. Yet dreadful. no two texts give exactly the same numbers Beyond the crudity of photolithography for the limits of this range in terms of wave- 256 Conveying Information about Color

length. Physicists, like the rest of us, have no color. I conclude that language is insufficiently clear conception of standard red or pure red. refined for some tasks or incapable of accomplishing them altogether. We need to seek a better understanding of those entities Color Notation and Ostensive that can only be explained by pointing at them. Definition Graphic symbols, including those used in rotational systems, are shorthand substitutes for Showing Colors words. Therefore, whether an abstract color We talk about things by mentioning their notational system can exist independent of names, the reason for complaints about the ostensive definition depends on whether colors ambiguity of color names. In an ideal language, can be described without resort to simile. The every object in the world would have a name, modern fashion among compilers of dictionaries and no object would have two names. This ideal is to identify green as a color like that of a is difficult to implement, for color or anything portion of the spectrum. The method is no more else. Although no two Americans have the same sophisticated than that of the proverbial Social Security number, we would be hard Ur-primitive who identifies green as the color of pressed if required to address other people by leaves. It is an inferior description, despite their Social Security numbers. Fewer names are resort to a more esoteric object. available than Social Security numbers, which Most leaves are green. A visitor from we tolerate in the assumption that the law of another galaxy could look at leaves to discover probability protects us. Undoubtedly an which color was meant. The spectrum is less excessively large number of Jane Smiths and helpful to the uninformed, because most John Smiths live in the United States. We trust portions of it are not green. An a priori that no single person will have to deal with understanding of what green looks like is several of them at the same time. necessary to determine which is the green Providing every citizen, among 200 million portion. For exactly this reason, nobody ever Americans, with a unique name would be a learned to identify green from a dictionary. formidable task. By estimate, only ten million If colors cannot be described in words or by colors exist, though the problem is similar. means of abstract notational systems Nobody would be able to remember ten independent of ostensive definition, this need million names, even if every individual not imply that color is shamefully tainted by variation of color had its own. the purely subjective. Its condition may parallel We manage to communicate about color, that of numbers and other conceptual despite these complications, because commu- primitives. Six, like red, has no scientific name nication is not limited to speech and writing. and is called six, written as six or 6, in Communication also encompasses a variety of technical treatises as in ordinary language. gestures, and human beings can communicate Similarly, no abstract notational system has by showing things to one another. Although been developed for up, down, left, and right. regarded less favorably than telling, showing is Like color and number, they can only be not a primitive ceremony. As is necessary in all explained ostensively. We cannot explain what modes of communication, common six means other than by displaying six objects, acculturation links the active subject (who just as we cannot explain the meaning of yel- shows) and the passive object (to whom some- low other than by exhibiting an item of that thing is shown). Rules determine what may be Conveying Information about Color 257 shown, to whom, when, and in what manner. Describing Colors Among objects that can be shown, color Describing the colors of objects, like showing swatches are displayed to make clear how a samples of the colors, follows everyday rules. particular color name ought to be understood In describing movable objects, we convention- or what color should be selected for a partic- ally tell what the object looks like, not where it ular purpose. Because the swatches convey is located. Describing, say, Picasso's Night information by requiring that something be Fishing at Antibes implies something other looked at, they resemble diagrams, charts, and than specifying, however meticulously, where graphs, specialized offshoots of written lan- the painting is situated relative to the elevators guage that are conventionalized and contex- at the Museum of Modern Art in New York tual. Graphs, for example, are not intended as City. The spatiotemporal coordinates of a port- representational pictures of what they "repre- able object, because subject to change, are not sent." A black bar can denote the aggregate of regarded as integral to the object. Yet identify- people who, according to context, purchased ing where an object is located can be legiti- new cars, live west of the Rocky Mountains, mate, sometimes uniquely for the circum- or are registered Democrats. stances. A trucking company sent to pick up Color swatches differ from diagrams, Night Fishing might be more interested in charts, and graphs in that what is shown con- locating the painting in the museum than in sist of colors rather than lines or shapes. The locating it within the history of art. At times, color swatch can be combined with other green paint is justifiably called the can on the forms, as in a layout for a two-color newspa- top shelf. per ad. The colors of swatches may be under- In daily affairs, colors are described by stood as samples, which implies they ought to methods not acknowledged in dictionaries. The be matched closely or exactly. Or, functioning colors of, say, house paints vary between more in the manner of a code, they can be manufacturers, even if labeled by the same taken to indicate only a broad class of color. name. If part of a room has been painted in Workers in a given industry, familiar with its Benjamin Moore ivory, and more paint is conventions, rarely find it necessary to discuss required, the hardware store clerk should be how a swatch, sketch, or layout ought to be told the brand name as well as the color. interpreted. Some of these ad hoc forms for identifying The textile manufacturer who wants to key colors reflect the built-in limits of manufactur- a dyed fabric to a painted piece of paper may ing processes. Photographic film varies in color find dyes are not available to match the colors sensitivity from one emulsion lot to the next, a exactly. A subjective determination is made of matter of importance to cinematographers. when a sufficiently close approximation has Knitting yarns vary in color from one dye lot to been achieved. Looser standards are involved the next. Specifying emulsion numbers and dye when a newspaper or magazine production lot numbers ensures color consistency between person, working from a red and black layout original and subsequent purchases. Use of lot of a two-color ad, "matches" the red of the numbers is common in industries that involve layout by specifying any available red ink. dyeing processes. The variables of dyeing, Color swatches and samples speak a silent lan- which include the length of time an object is guage of their own. Their colors are meant to immersed, are such that color consistency often be matched. But how closely depends on the cannot be guaranteed from one batch to the circumstances. next. 258 Conveying Information about Color

An emulsion number printed on a box of dark bluish red, and so forth. movie film is a code. It tells nothing about the Dictionaries rarely acknowledge opera- color sensitivity of the film other than that it is tional usage, which bypasses ordinary rules. In similar to that of other boxes of film bearing its formal meaning, a chemical name such as the same number. The emulsion number does zinc sulfide identifies a substance, not its color. not, interestingly, guarantee that the film is But scientific terms often drift into ordinary properly balanced for color sensitivity, and it language by acquiring everyday meanings may warn that it is not. Boxes of film may inconsistent with technical usage. These mean- carry a notation, say, that all rolls with emul- ings can identify how an item is used, rather sion number 1,174 should be used with a than what it is. Fluorides become popularly filter of a particular density. known as ingredients in toothpastes; bromides, aids to digestion; chlorine, a killer of germs in swimming pools; iodine, a household Chemical Names as Color Names antiseptic. Codes and names are not always easy to sepa- No chemistry book would define hydrogen rate, because names can be used as codes. as a substance used in bombs or neon as a gas Many commercial color names for artists' used in neon signs (which in most cases are paints, say, cadmium red, are adapted from the filled with gases other than neon). By a simi- chemical names of the pigments used in the larly operational rationale, the chemical names paints. Cadmium red is a red paint in which the of pigments used to apply color to surfaces pigments are cadmium sulphide and cadmium acquire a secondary meaning as the names of selenide. Pigments are a specialized type of those colors. object, and the practice of naming colors after When chemical names acquire a secondary objects is commonplace. meaning as color names, different rules apply For a chemist, the difference between zinc than when the names are used within the sulfide and cadmium sulfate might turn on the parameters of chemical nomenclature. Zinc difference between a sulfide and a sulphate. yellow paint from two different manufacturers When chemical names are adapted to identify is likely to differ in color, method of manufac- pigments, a more important consideration is ture, and such variables as ratio of pigment to that zinc sulfide (zinc yellow) is a pale green- vehicle. Yet both will be called zinc yellow. ish yellow pigment or color. Cadmium sulfate By extension, the name can be applied to any (cadmium yellow) is also yellow, but brighter yellow that looks similar to zinc yellow, and not so greenish. whether or not attributable to that pigment. Although many names for artists' pigments We need not investigate the chemical compo- are borrowed from chemical nomenclature, sition of the glaze used for dinner dishes their use is not consistent with that nomencla- before describing the color of the dishes as ture. The chemical names acquire an opera- zinc yellow, cobalt blue, or cadmium red. The tional meaning as the names of colors. A piece chemist, grading a batch of zinc sulfide, finds of fabric described as cadmium red is under- chemical purity, an absence of adulterants, sig- stood to be a particular color. We are not nificant. For the paint manufacturer, the pri- required to assume that cadmium salts were mary question is whether adulterants affect used to dye the fabric. For anyone familiar color. with names of pigments, cadmium red dark Patterns emerge from operational usage. conveys more information about the intended The names of modern dyes (mauve is an color than alternatives such as red, dark red, exception) seldom inspire color names. Many Conveying Information about Color 259 of these dyes have no ordinary language names color cataloging that relies on number alone. In and are identified by chemical names that are an entirely numerical system, each of the ten polysyllabic and difficult to remember. Color million ISCC colors might be assigned a names adapted from the chemical names of number between one and ten million. In theory, pigments refer to those of modern vintage, we could then include in ordinary language including cadmium red, cadmium yellow, and statements such as "I like color number titanium white. Traditional pigments tend to 965,873." Assuming that the ISCC is correct in retain their original names: vermilion, emerald computing ten million visually distinguishable green, burnt sienna, yellow ocher, and so forth. colors, no ambiguity would occur about which color is meant. The barrier is the limitation of human memory, considerable in this case. Human beings are unable to commit colors Colors and Numbers to, or retrieve then from, memory with Most modern color cataloging systems use exactitude. This is why the wise, when asked to numbers as well as names. In the ISCC-NBS decide whether two colors match, insist on system, ten million colors are divided among looking at both colors at the same time, in close 267 classes, while finer distinctions are proximity to one another. We lack a memory for indicated numerically. The English language color and also lack the computer's ability to does not contain ten million color names. If it manipulate long strings of numbers without did, nobody would be able to memorize them. error. Although my friends cannot be expected The advantage of numbers is that anyone who to remember my Social Security number, the understands the system can extend it proliferating computers in which it is entered indefinitely. Counting to ten million does not always print out the digits correctly. In the case require the prior memorizing of ten million of color, nobody would remember, or be able to numbers. visualize, the difference between, say, color Like the ISCC notational system, Munsell's number 965,873 and color number 965,874. Yet combines names with numbers, though in a the difference, if between ISCC colors, would different manner. Hue (major color class) is be noticeable when swatches were examined. described by name, say, red. For notational Modifications to numerical systems can purposes, each hue name is abbreviated to a make them easier to remember. The following letter (R, in the case of red). Value and set might be used as the basis for a system of chromaticity are indicated by numbers. The net coding the spectral hues. effect is to address the problem that any given color name-for example, light gray greencan be shown to be applicable to a broad range of colors. Particular colors or shades of color in the range can be efficiently designated by any of a wide variety of name and number combinations, by such constructions as light gray green 1, light gray green 2, light gray green 3, and so forth. The device eliminates any necessity for coining individual names for each shade or variety of light gray green. Although numbers are often used as quail- In this system, a blue-violet, intermediary fiers in this manner, I know of no system of between blue (code 2) and violet (code 1), 260 Conveying Information about Color

might be designated by an intermediary number ratio between their respective assigned carried out to any required number of decimal wavelengths. places, for example, 1.5. The terms 1.1, 1.13, Munsell's notation for what he called a 1.28, 1.47 might designate a series of typical maroon is SR3/4. R stands for red. 5, 3, blue-violets that tend increasingly toward blue, and 4 refer to hue, chroma, and value for that a system that ought to be workable whether the maroon. Less ornate constructions are possible. number of colors is infinite or not. The system Typical maroon can as easily be identified as resembles a clock, which enables us to divide 534, 435, or 999. The sole requirement for an time into hours, minutes, seconds, and other internally consistent code is that a single other neat packets, though we believe time itself number apply to no more than one color. What is a continuum and possibly infinite. offends in identifying colors by numbers are We easily lose track of the passage of time, meaningless numbers, which code without the reasons watches and clocks are needed to purporting to measure. If 5 is used as a issue reminders. No reminder is needed of designation for orange, or 534 for maroon, the what red looks like. Color is a more intimate numbers are as arbitrary as those in the Dewey concept than time, more closely tied to decimal system used for library cataloging. perceptual experience. Perhaps for this reason, a They are also potentially as useful-and less sense of discomfort interjects itself when 6/4/2 ponderous than the systems at hand. or any other arbitrary group of numbers is Among available color atlases, the Munsell proposed as an alternate set of names for Book of Color contains over twelve hundred red/yellow/blue. One argument against an different swatches. Its utility is not impaired if entirely numerical system for coding colors is as each swatch is renumbered according to follows. position in the book. Whether the numbering The number continuum cannot reflect the commences at the first or last page is of no differences in color quality between different importance. A designation such as 534 is as sectors of the spectrum. Blue is fundamentally useful as 583/4. It may even be more helpful, dissimilar to orange. This is not made clear if because it does not create the misleading blue is called 2 and orange is called 5. Numbers impression that something measurable has been are insufficiently expressive. The visual measured. difference between, say, blue and orange is Thinking of orange as, say, 5 seems what they fail to express or communicate. strange. But this is just a matter of The demand for expressiveness through acculturation. Arbitrary numbering systems are names is not consistent. Number is apparently used to describe perceptual continua in which acceptable when something is thought to have sectors differ as greatly as orange and blue. The been measured. The prime example is the thermometer measures the temperature defining of red, in physics, as light with a continuum in terms of arbitrary degrees. wavelength of 650 millimicrons, violet with Understanding what the numbers on a 400 millimicrons, and so forth. The thermometer mean implies a complex quantitative relationship that exists between the interpretive process. Changing from one frequencies raises the question of whether a method of measuring temperature to another, parallel quantitative relationship can be say, Fahrenheit to Celsius, becomes confusing assumed between the colors. Probably it cannot because a new system of interpretation must be be, or the assumption is not useful. There is no superimposed on an old one. We eventually way in which violet is approximately learn to remember that 100°C is hotter than two-thirds of red, other than that this is the 200°F and 0°C is warmer than 20°F. Conveying Information about Color 261

For those familiar with the conventions, Measuring differs from counting. Often no -20°F means unbearably cold; 125°F means operational criteria can be established to unbearably hot. In-between numbers indicate explain what twice as much means. It can conditions that are called cool, bracing, cold, mean twice as large a measurement on an pleasantly warm, unpleasantly warm, very hot, arbitrary scale and nothing further. The reading or chilly. We become adept at interpreting the from the scale fails to correspond to any numbers and know what they imply. identifiable aspect of the object and is less a Temperature readings, in turn, impress by their measurement than an ornate code. seeming exactness, and may people prefer According to the Fahrenheit thermometer, them to verbal description. To hear that the the temperature at which water boils is 65/s temperature rose to 90° seems to communicate times as high as the temperature at which it more than just being told it is hot. If colors freezes. But 212 °F cannot be 65/s times as hot were coded numerically, people would as 32 °F in any demonstrable sense. The ratio similarly learn to interpret the numbers or fails to hold if readings for boiling and freezing would come to regard them as meaningful. water are taken from Celsius thermometers. A system that measured some aspect of When readings are taken from arbitrary scales, color (and reported the measurements) might the readings cannot be regarded as seem preferable to one in which arbitrary measurement, and reasoning based on numbers were assigned. When numbers numerical ratios is meaningless. represent measurements, they "really mean In measurement of another type, something something," in that they communicate can be perceptually verified. That one box information beyond themselves. If a person is contains twice as many apples as another can six feet tall, the measurement conveys be factual, verifiable by counting the apples in information about the individual: that he or she the boxes. A ten-pound box of apples can be is tall. Except to personnel in the Social shown to weigh as much as two five-pound Security system, that the six-foot tall person's boxes. But to say that one color is twice as dark, Social Security number is 683-97-0014 or twice as blue, as another is arbitrary. No way conveys nothing except the information that exists of verifying the statement or explaining this is the number. what it means. Similarly, a temperature of 80° We have many systems that purport to cannot be shown to be twice as hot as a measure color or some aspect of it. But the temperature of 40°. I conclude the reason for urge to measure is not sufficient to ensure that the difference is that apples can be isolated for measurement really takes place, or that the counting. Darkness, blueness, and warmth numbers are interpreted in a sensible manner. cannot be isolated as objects, which bears on Munsell grades the grays on a value scale the limited nature of the measurements we ranging from 0 (black) to 10 (white). This scale make of them. represents "the averaged results obtained by A main argument for coding colors by seven experienced observers" (Munsell f 1905] assigning arbitrary numbers to them is that the 1961, 63). On this scale, gray 4 is darker than alternative of measuring (assigning meaningful gray 8. Whether twice as dark is a meaningless numbers) may not be available. Some sort of question, because the ratio between the Dewey decimal system for colors might be respective grays (8/4 -2) points to nothing that preferable to arbitrary numbers disguised as is perceptually verifiable. Calling one color measurements of hue, value, chroma, or twice as dark as another is as meaningless as whatnot. Orange as 534.7 is ultimately less calling one day twice as hot as another. objectionable than SR3/4. CHAPTER 29 On Ambiguity in Color Names

It must not be supposed for a moment that the colors on this chart represent the colors of all the birds of eastern North America . . . . It should be clearly understood, therefore, that when greyish brown, for example, is mentioned, it does not follow that the feathers to which the term is applied are of exactly the same color as the plate, but that they are nearer to this color than to any other in the plate. Used even in this general way, the plate will prove a far more definite basis for description than if everyone were left to form his own idea of the colors named. Frank M. Chapman, Handbook of Birds of Eastern North America

lorence Elizabeth Wallace compiled color name, but its nearest equivalent in En- many examples of failure of parallelism glish, warm colors, is not, translations are bet ween color names in English and confusing and misleading. those used by ancient Greek authors (Wallace Failure of parallelism between languages 1927). Homer had no label for the color mod- cannot fully account for the frequency with ern speakers of English call bright yellow. The which color names are mistranslated, poorly Greeks regarded orange as three classes of hue, defined, inconsistently used, and misused in corresponding to our red-orange, orange, and English by great translators, great scientists, yellow-orange. Although common between compilers of dictionaries, and educated adults. one language and another, skewings of this Although taught to choose words carefully, we type create difficulties for translators. Bright are careless with color names. Nobody makes yellow cannot be rendered in classical Greek if the mistake of identifying pink as green. Pri- this variety of hue has no name in that lan- mal color names are usually remembered, as guage. The Bassa color name ziza cannot be are the more familiar symbolic associations translated into a single nontechnical English with these colors. Misused or confused names word if we want to avoid warm colors and the are usually those of the second echelon, fami- class to which ziza refers includes reds, liar enough to be remembered but not familiar oranges, and yellows. Alternately, if ziza is a enough to be remembered entirely, say,

262 On Ambiguity in Color Names 263

magenta, vermilion, scarlet, lavender, crimson, term suggests, redundantly, "more fiery than or viridian. most (fiery) reds." As if colors were not worth Color names of this kind are not obscure concerning ourselves about, we use color words. Most are sufficiently familiar to be names in a haphazard manner. Dictionaries list appropriately included in a spelling bee for many scientific names and place names but are children. An educated adult can spell them and less comprehensive with color names, the has a general impression of what each means, reason specialized dictionaries of color names though they remain capable of causing uncer- have been compiled. Even scientific theories tainty. Consider the Italian word vermiglia in about color tend to be more slapdash, less several translations of Dante's Commedia. Ver- carefully considered, than theories of other miglia is identified as a cognate for the English kinds. vermilion in many Italian-English dictionaries. Some color names are inherently vague. Cognates, where they exist, are the preferred With the best will in the world, nobody could translations for foreign language terms, unless learn how to use them. Violet, purple, and lav- a word changes meaning in passing from one ender are not regarded as synonyms. Each language to another. Yet in major translations refers to a different color, intermediary between of Commedia, vermiglia is often not translated red and blue. Ranking and range are uncertain. as vermilion. It is not even translated to the Is violet more bluish than purple? Is purple same English color name in each occurrence. more reddish than lavender? Is purple a variety In the Temple Classics edition of Inferno, of violet? Alternately, is violet a variety of vermiglia is crimson in one passage, red in a purple? second, and fiery red in a third (Inferno iii, Answers vary among those imprudent 132; Inferno vi, 16; Inferno xxxiv, 39). John enough to risk opinions. For Wallace, violet Ciardi translated vermiglia as red in the first meant bluish purple and was not the correct two passages and as fiery red in the third English word for translating two Greek color (Inferno xxxiv, 39). In the Reverend Henry F. names used by Homer that refer to reddish Cary's translation of Inferno (1805), vermiglia purples (Wallace 1927, 74). Answers are not is vermilion in the first and third of the that simple. In English, consensus cannot even passages but red in the second (Inferno vi, 16). be found about whether the color at the short Ezra Pound and others have written about extreme of the spectrum is more properly called the philosophy of translation as it applies to purple or violet. poetry. I suspect that Pound's erudite The human eye shows greatest sensitivity considerations are not the issue in this case. to fine differences of color in the yellow-green Dante is highly effective in his use of color range. Yet second-echelon color names for names. Assuming that he had the same color in reds and purples outnumber those for yellow- mind each time he used vermiglia, we must green or any other range of color. Examples look to the translators for an understanding of include maroon, scarlet, crimson, vermilion, the inconsistencies. They were uncomfortable cochineal, magenta, cerise, heliotrope, violet, with the color name vermilion or thought that lavender, purple, fuchsia, and rose. Why do their readers would be. we have so many names for reds and purples? The lack of consistency or refinement in Does a cultural preference for those colors the substitutions is a more subtle defect. exist? Among terms selected,crimson and red are not A cultural aversion is more likely. The synonyms for vermilion. Fiery red is without colors inspire apprehension, which does not grace, because all bright reds are fiery. The explain why this should be associated with a 264 On Ambiguity in Color Names

more-than-usual number of names. Conven- example, encounters this difficulty with several tional canons of taste dictate that color schemes color names of the second echelon. These are dull if limited largely to brown, gray, blue, include magenta, fuchsia, blue, indigo, and and green. More definite disapproval is reserved violet, a cluster of overlapping names for for uses of red and purple considered excessive, related ranges. excessive because too large an area of the color Magenta is defined as (1) the dye fuchsin, is used or the shade is too bright. Too much red and (2) a reddish purple color. The expectation is considered vulgar. Too much purple is is that fuchsin will be properly identified, in its thought bizarre. The association of purple with place, as reddish purple (magenta) in color. My homosexuality (it less often carries, today, dictionary's entry for fuchsin calls it a dye that connotations of royal robes of Tyrian purple) is forms deep red solutions and derives its name incidental to its being considered (as is orange) from the fuchsia flower, which it resembles in inappropriate for many purposes, ranging from color. Is the fuchsia flower (resembling fuchsin men's business suits to bathtubs to wallpaper. in color) magenta, reddish purple, or deep red? Names can be regarded as surrogates for the The entry for fuchsia identifies it as the name of named object, the reason human beings dislike a crimson flower. having their names mispronounced, misspelled, The reader is left to puzzle over whether or forgotten. They read the lack of attention as magenta means reddish purple, deep red (like an insult, and a common way of devaluing the dye fuchsin, with which the color is anyone or anything is to pay no attention to the identified), or crimson (like the fuchsia flower name. The more-than-usual ambiguity in names that fuchsin resembles in color). Greater tangles for purples is undoubtedly a side effect of are avoided only because this edition of the societal ambivalence toward the color over a dictionary declines to hold forth on what period of centuries. People who like purple fuchsia means when used (as it often is) as a clothing often say they like the color because it color name. is "different." Orange, regarded by the Greeks as three colors, has a controversial history and is said to Blue, Indigo, Violet be disliked in Japan. Munsell, arguing that My American College Dictionary defines orange is not a color, rechristened it yellowred. spectrum by listing the colors as Newton Ostwald restored it as orange, whether or not identified them. The last three are blue, indigo, following Moses Harris, who had defended the and violet. The entry for indigo calls it a deep integrity of orange centuries earlier. violetblue between blue and violet in the spectrum. Among blue, indigo, and violet, the reader is to understand that violet is the least bluish (most reddish) color. The compilers have Magenta and Fuchsia forgotten this ranking in a third entry, which The greater number of color names in English says that violet means bluish purple. A fourth are borrowed from object names, from objects entry explains purple as any color containing of that particular color. The system sounds red and blue. simple, but leaves room for confusion about Munsell, who regarded purple as a "more which object ought to be pointed out as an appropriate" name for violet, would have example of a particular color. Compilers of dic- objected to the third and fourth entries. They tionaries become as confused as anyone else. imply that in the color class purple the more My American College Dictionary (1953), for bluish ( not the more reddish ) varieties are On Ambiguity in Color Names 265

called violet. The dictionary identifies the last both sexes, except that the vendor had no pink color in the spectrum as violet. Munsell called it shirts or lavender shirts for men. Readers were purple. Newton, followed by twentieth- expected to understand why. With a "dare to be century colorimetrists, reserved the name purple different" attitude, Brooks , the for mixtures of red and violet, the extremes of department store, once offered pink Oxford the spectrum. In Newton's arrangement purples button-down collar shirts for men. Pink, I were always more reddish than violets. conclude, is not a color for men. Even in this era My dictionary parts company with Munsell of androgynous rock stars and unisex clothing, in indicating that lavender, violet, and mauve pink is associated with baby girls. But if a pink are bluish purples, and lilac is reddish purple. shirt comes from a fashionable haberdasher, the Munsell characterized each of these colors as ordinary rules do not hold. a purple that is neither reddish nor bluish Pink button-down shirts from Brooks (Munsell [ 1905 ] 1961, 56). The few certain- Brothers were popular for many years. Violet ties in this morass are not those identified by shirts were not offered. A timid person is called my dictionary. Purple and violet, as Munsell a shrinking violet. Pale shades of violet and argued, are virtual synonyms. People use lavender were once thought to be appropriate whichever they prefer for the name of, say, the colors for clothing for elderly women. Market last color of the spectrum. Violet names a researchers find, I am sure, that violet shirts for broader range of color than that of the flower. men do not sell well. These ideas survive What color names mean, how people use because people sense an inner logic, often and misuse them, and which colors are without knowing how to explain the logic. Taste preferred is an aspect of popular culture. Many in color, like taste generally, grows from a set of people have favorite colors and color expectations. combinations. They regard the preferences as personal taste. But many ideas about the use of color are traditional and crop up again and Orange, Purple, Yellow-Green again. Black and white color schemes, we say, We expect the colors of the spectrum to be are improved by a little red. Yet too much red "pure," which means both bright and unique. is garish or in poor taste. The color uniqueness of red, yellow, blue, and Dressing entirely in red is thought to convey green is rarely questioned. Orange is said to a message of aggression, though only in certain resemble red, which may cast aspersions on its contexts. British soldiers wore red uniforms in purity. Purple is said to resemble blue. Along the 1700s and became known as redcoats. with bright yellow-green, bright orange and Modern male corporate raiders and real estate purple are not ordinarily used as a basis for developers, no matter how aggressive, do not decorative schemes, whether in room interiors wear red suits. A man's red suit might be or clothing. The combination appears in Art thought acceptable for an artist, actor, singer, or Nouveau and other styles that aspire to be anyone in an occupation where standards are different. thought to be different. A vase of red roses is Yellow-green, like yellow-orange, is prob- not thought of as aggressive. Red sports cars are ably regarded as an impure yellow, less than thought of as daring. Black sedans are regarded unique because it resembles another color. In as conservative. its muted form, as olive drab or , yellow- I received a mail-order catalog offering green is perennially fashionable, as is rust, a shirts for men and women in a choice of thirty muted form of orange. Bright yellow-green is different colors. All colors were offered for disliked and regarded as unflattering in cloth- 266 On Ambiguity in Color Names

ing. Few names exist in En- wald's blue that the two could never be taken glish for colors in the yellow-green range, far one for the other. fewer than for reds and purples. Yet yellow- Beginning from this difference about what green is the range in which the eye shows constitutes blue, each theorist presents a system greatest acuity. for naming the major hues grievously skewed in comparison to the system of the other. Munsell includes ten colors in his wheel. Ostwald Identifying Colors on Color Wheels switches to eight, incommensurable with ten. Anthropologists and linguists have long been Newton, on numerological reasoning, had interested in where the points of division identified seven. Moses Harris preferred six; between colors of the spectrum are imagined Johannes Itten, twelve. The question is how to lie. How colors are named bears on the many ways a pie can be sliced. The Rational question of how acculturation affects interpre- Color Circle, published by Faber Bitten in tation of visual experience. V. F. Ray, listing 1934, is divided into thirteen slices and is the used in ten North Ameri- asymmetric. The slices radiate from a point can cultures, found little uniformity in the halfway between center and edge, rather than terms used or the parts of the spectrum to from the center (Birren 1969, 25). which they apply (Segall, Campbell and Her- In Munsell's system for naming his ten skovits 1966, 47; Ray 1953). The point can be hues, red, orange (which is called yellow-red), made without investigating whether any ten and yellow occupy 30 percent of the wheel and American Indian tribes name the major colors by implication constitute 30 percent of the in the same way or whether we do it as the spectrum. In Ostwald's system, these colors are classical Greeks did. No two modern experts allotted 37.5 percent, an amount 25 percent on color agree with one another, even on larger than that allowed by Munsell. Two of issues as fundamental as the names of the Ostwald's color classes, leaf green and purple, major classes of color, the names of the colors together occupy approximately 25 percent of of the spectrum, or the way hues should be his color circle. Munsell's corresponding arranged on color wheels. colors, yellow and green-yellow (the range On Albert H. Munsell's wheel, the colors corresponding to leaf green) and purple and of the spectrum are arranged to read clockwise red-purple (corresponding to purple) occupy from the color of longest wavelength (red) to 40 percent of the total color circle, an amount the shortest (violet). Perhaps to stake out a 60 percent larger than that allocated by claim to product uniqueness, Wilhelm Ost- Ostwald. wald's is designed to read counterclockwise, Ostwald must have believed that Munsell moving, again, from red to violet (Munsell underestimated the number of discernible vari- [ 1905 ] 1961; Jacobson 1948, 26; Munsell ations of color in the range identified as leaf 1969, 82; Ostwald [ 1916] 1969, 83). The green to purple. But Ostwald's reasoning goes color Ostwald identified as blue is the range unexplained. The confusion caused by the popularly called royal blue. Munsell included skewing is worsened by inconsistency in nam- a similar color among the major hues but called ing. Munsell and Ostwald were not of a single it purple-blue; it is not, in other words, blue. mind about the qualities that make a color Munsell's blue, more toward green, looks like name appropriate. Munsell, who "decried such the color Ostwald called turquoise. Thus, Ost- vague terms as pea green, evergreen, invisible wald names a royal blue as blue; Munsell green" (Munsell 1969, 71), left no message names a turquoise blue so different from Ost- about whether he also decried sea green and On Ambiguity in Color Names 267 leaf green, which in Ostwald's system name As the Munsell and Ostwald systems stand, two of the eight major hues. Sea green is a large numbers of statements about particular traditional name, rather than a coinage by Ost- colors can be constructed that are true according wald or his translators. The cube of colours by to one system but not the other. And any William Benson includes the names sea-green naming of colors correct according to the green and sea-green blue (Benson [ 1868] Munsell system will be incorrect according to 1930). No plain green occurs in the Ostwald Ostwald. These ornate systems, incompatible system. Hopefully, intermediaries between sea with one another, are similarly incompatible green and leaf green are not to be designated with less fully developed systems suggested by by such constructions as sea greenish leaf Newton, Moses Harris, Johannes Itten, Faber green, or leaf greenish sea green. Birren, and others.

PART FIVE Color Theory

Various contrivances have been proposed under the titles of Tables, Scales, Colour-Circles, Chromatometers &c., for representing either by numbers or a rational nomenclature, colors and their modifications. They are generally founded on these three propositions:-I. There are three primary colors. 2. Equal portions of these colours being mixed, produce pure secondary colours. 3. Equal portions of the three primary colours produce black. But, I. We know of no substance which exhibits pure colour; that is, which reflects only one kind of coloured rays, whether pure red, pure yellow, or pure blue. 2. Since it is impossible to procure pure colouring matters, how can it be said that orange, green, and violet are composed of two simple colours mixed in equal proportions? Or that black consists of a mixture of equal parts of three simple colours? M. E. Chevreul, The Principles of Harmony and Contrast of Colors

CHAPTER 30 Systematizers and Systems

The earlier assumption of six principal colors and three fundamental colors is incorrect. Wilhelm Ostwald, The Color Primer

he invention of the color wheel is him to conclude that the planets, the musical usually attributed to Newton, though notes, and the hues of the spectrum must all occasionally to the painter Moses match in number and had all been created in Harris (ca. 1766). We do not know who origi- sets of seven. nated the idea of arraying all the colors in a Mathematical conceptions of order con- three-dimensional color solid, an activity tinue to hold sway in the natural sciences. But popular from the seventeenth century onward. the computations have grown too complex to be Relationships among colors were diagrammed understandable to a wide audience, and the on triangles, spheres, cones, cubes, and other practice of assuming simple correlations has geometrical shapes. long fallen out of favor. The search for simple The diagrams are usually symmetrical, measures continued longest among those who reflecting dreams of a neatly ordered universe believed that the arts followed laws comparable that offers mute testimony to the tidiness of to those of the sciences. A search for morality God's housekeeping. Theorists assumed that was the hidden agenda, a determination to natural relationships, including those among define beauty and stamp out the unbeautiful. colors, could be reduced to mathematically Well into the twentieth century, many seekers regular measures. The tone had been set by appeared to believe that the arts, especially Newton, whose numerological interests led music and the visual arts, could not be con-

271 272 Systematizers and Systems

sidered genuinely respectable unless laws could one apex, white at the other; Charles Blanc (ca. be found that explained them and defined their 1873), a six-pointed star. A. Hofler (1905) purposes. diagrammed a double pyramid with black, If color harmony and other forms of beauty white, red, yellow, green, and blue at its could be shown to have a rational basis, the arts vertices. Other mappings came from would be revealed as more than just aesthetic. Charpentier (1885) and from Athanasius They would assume a position among the Kirchner (1671), the German Jesuit better verities, an expression of natural law. Only then known for his interest in mechanical methods of could they be thought to be touched by the musical composition. moral (or perhaps masculine) nature of the The search for symmetry has parallels in Creator, who ordained order by bestowing the other areas of design, the visual arts, and music. laws by which creatures and phenomena were Jay Hambidge's dynamic symmetry, based on bound. measurements of the Parthenon, gave The ceremonial sense of high ethical prescriptions for inherently beautiful concern was pronounced by the early 1900s, proportions, the proverbial lost formulas of the more in theories about the mathematical basis ancients. Joseph Schillinger (d. 1943), professor of color harmony than in application of similar of mathematics at , reasoning to other areas of the arts and attempted to reduce beauty, and through it, the aesthetics. Color theory, in the popular mind, arts, to mathematics. Schillinger said George remains largely chained to these fancies. We do Gershwin composed Porgy and Bess according not know, however, whether natural law has a to the Schillinger system of musical mathematical or symmetrical basis, whether composition (Schillinger 1941; 1948). We tidiness has a superior claim to moral virtue, or know it was used by Heitor Villa-Lobos, whose whether the universe is tidy. New York Set to Music (1946) was advertised Alfred Hickethier's by as a translation of the New York skyline "from Numbers includes illustrations of eighteenth- silhouette to music notes with the help of the and nineteenth-century color diagrams Schillinger system of Musical Composition." In (Hickethier 1963, 15-17). The German Mathematical Basis of the Arts, Schillinger astronomer Tobias Mayer (1723-62) developed offers formulas for devising harmonious color a three-part of earlier vintage schemes. The formulas emphasize regularity than the Maxwell triangle adapted by modern but otherwise bear little resemblance to earlier colorimetrists. The mathematician J. H. recipes by Munsell and Ostwald. Lambert (1728-77) invented a color pyramid. Whether color harmony and beauty of Goethe, who conducted color experiments for proportion follow mathematical laws is twenty years, drew a variety of color circles uncertain, and the terms are difficult to define. and triangles. The Romantic painter Philip Otto The proportions of the golden section have Runge (1777-1810) developed a color sphere been called beautiful but would become boring adapted by Albert H. Munsell and later by if we used them for everything. Arguing in Wilhelm Ostwald. In Runge's sphere the hues favor of an association between mathematics lie at the equator, black and white at the poles. and music is easier, because ratios exist among M.E. Chevreul (1786-1889) arranged the colors vibrating strings. Louis Wilson's ophthalmic on a hemisphere and the hues in a circle. color scale is based on an assumption that rules Wilhelm von Bezold (1876) put the colors in a for musical harmony can be directly adapted cone grading toward a black apex. Ogden Rood to create pleasing color combinations (see fig- ( 1831-1902 ) devised a double cone, black at ure 34-1). Indeed, color harmony is a concept Systematizers and Systems 273

of musical origin, as are the chords and Systems for selecting harmonious colors intervals mentioned in the writings of Munsell, never sank entirely to the vulgarity of forgotten Ostwald, Bitten, and other color theorists. schemes for creating musical masterpieces. The Analogies between colors and musical emergence of color styling as a profession notes are vague, and Arnheim questioned their provided an incentive for minimizing claims relevance (Arnheim 1956, 338). Benjamin that no sensibility to color is needed if the right Whorf criticized "what is often rather inac- recipes for using it are at hand. Ratios curately called the `music' of words," as well comparable to those between musical notes do as synesthesia, the process by which "we speak not exist between colors or among their of `tones' of color, a gray `monotone,' a `loud' wavelengths. In musical notation, B-flat is a necktie . . . . Colors are conjoined with feelings single note. Red is a range of thousands of for the analogy to concords and discords." colors. If red is to be explained in terms of Whorf called for a more refined aesthetic wavelengths, many wavelengths are involved. sense, an end to synesthetic "confusion of Classic works in the history of color theory thought" (Whorf [ 1956] 1967, 156, 267). include Boyle's Considerations Touching Systems for composing music, like those Colours (1664), Newton's Opticks (1704), for combining colors, were perennially con- Goethe's Farbenlehre (1810), and Chevreul's De troversial. Well before the twentieth century, la loi du contraste simultane des couleurs however, they boasted of practical benefits. (1845). None was widely circulated among a The late-eighteenth-century London music popular audience, and the issues they raise are publisher Weleker announced a "tabular technical and complex. The most popularly system whereby any person, without the least influential and commercially successful of knowledge of music, may compose ten modern theorists were Albert H. Munsell thousand minuets in the most pleasing and (1858-1918), Wilhelm Ostwald (1853-1932), correct manner." Martin Gardner (1974) traced and their interpreters including Faber Bitten. the steps by which this interest led to the Although Munsell was an art educator and question of whether a computer can be Ostwald a physical chemist, their reasoning programmed to compose symphonies without about color was similar. Each conveyed a human intervention. simple message carried directly to the Is the core of the arts human intelligence, marketplace. The two men met in Boston in as Piet Mondrian thought was the case with the 1905, that fruitful year in which Picasso made art of painting? Can a chimpanzee, pounding the first Cubist paintings and Einstein and Freud randomly on a typewriter, eventually compose began publishing their ideas.' Milton's Paradise Lost? We less often ask Munsell, educated at the Massachusetts whether the chimpanzee, given a brokerage Normal Art School, taught at the Normal Art account and a pin to stick randomly in the stock School in Boston until 1915. In 1879, he read market page, will eventually make a killing on Ogden Rood's Modern Chromatics (1871), a Wall Street. In the visual arts, paint-by- popularization of Maxwell's ideas. Rood's book the-number sets and coloring books speak to was read by the French Impressionist painters, the yearning for easy fixes among a popular along with Chevreul's. audience, as do books on how to get rich. In A Color Notation (1905), Munsell out- Modern coloring books for grown-ups project lined his system for regularizing the under- undertones of irony, as if a recognition that standing of color and ensuring good taste in filling in blanks, systematically or not, is a sim- its use. The system was later expanded by the pleminded activity for an adult. Inter-Society Color Council ( National Bureau 274 Systematizes and Systems

of Standards). Dorothy Nickerson, Deane B. which still actively merchandises industrial and Judd, Kenneth L. Kelly, and others who worked pedagogical materials. Bitten recommends, as on the ISCC-NBS system have also been "of particular interest and value . . . certain contributors and consultants to later editions of teaching aides offered by the Munsell Color Color Notation. Company. These are reasonable in cost and Ostwald, born of German parents in Latvia, have been widely used in color and art studied at the University of Dorpat (Estonia) education. Available are a number of charts and and held the post of professor of physical chart sets on which color chips may be chemistry at the University of Leipzig mounted. There is a hue, value/chroma chart (1887-1906). Retirement to Saxony enabled showing Munsell's five principal colors and five him to devote his time to the study of color and intermediate 'colors, a nine-step gray scale, and natural philosophy. The first of Ostwald's a seven-step chroma scale for red. In addition several books on color, Die Farbenfibel (Color there are sets of constant hue charts for Primer) was published in 1916. Munsell's key hue .... There is a teacher's The Munsell and Ostwald systems are demonstration kit and special large wall charts" incompatible. We cannot take the recipes for (Munsell 1969, 796). harmonious color schemes offered by one sys- Ostwald's corresponding venture into tem and apply them to the other. But both sys- industry, including the education industry, was tems follow the same format. Each offers a gray "known as `Ostwald Energie,' and this scale displaying a series of steps between white organization not only reached the entire and black, a hue circle identifying the major German educational system but offered to spectral hues, a color solid that theoretically extend counsel to German business and can be expanded to include all colors except the industry" (Ostwald [ 1916] 1969, 5). Ostwald's metallics, and a color manual that displays the Color Science was translated into English by J. colors of the solid in book form. Each of these Scott Taylor. More widely circulated in the items is presented in the form of paper United States than Taylor's two-volume standards: colored swatches in various arrays. translation are adaptations of Ostwald's ideas by Both theorists regulated the naming of Egbert Jacobson (1942, 1948), Faber Bitten colors, developed systems for writing color (1969), Charles N. Smith (1965), Rudolph notations, and set forth formulas for creating Arnheim (1956, 335-38), and others .2 Egbert harmonious color schemes. A different set of Jacobson, formerly an art director for Container parameters is used in each system. Munsell Corporation of America, supervised that located colors according to the coordinates of company's publication of The Color Harmony hue, value, and chroma. The Ostwald Manual, a set of standards based on Ostwald's coordinates are hue content, black content, and theories. Jacobson's major work in his own white content. Ostwald's system, patterned after name is Basic Color (1948), an interpretation of Munsell's also drew on the theories of Gustav the Ostwald system showing ways of applying Fechner. The logarithmic gray scale, one of it to industrial styling and the fine arts. Bitten, a Ostwald's applications of Fechner's ideas, has color consultant and founder of Faber Birren been incorporated into recent editions of and Company, was born in Chicago (1900) and Munsell materials. studied color theory with Walter Sargent in the Both theorists founded companies that dis- School of Education at the University of seminated their ideas to a wide audience. In Chicago (Sargent [ 1923 ] 1966). While 1918, the year of his death, Munsell organized Jacobson is an interpreter of the Ostwald sys- the Munsell Color Company ( Baltimore ), tem, Birren is the system's exegete, an author Systematizers and Systems 275

of many books on color and color systems.3 using color, rather than its abstractness, Creating harmonies by means of the Ost- complained that monotony in the spacing of wald system is simple, outlined by Jacobson intervals in the color solid implies monotony of in Basic Color. Familiarity with the arrange- colors (Sweeney 1945, 34). Matisse, a major ment of the color solid is required. Sets of colorist, said he devised his color schemes colors are selected from the solid according to intuitively, and denied using a system. Van formulas Ostwald prescribes. Any set chosen Gogh, said to have chosen the colors for his by this technique is harmonious; ostensibly it is paintings by playing with scraps of colored more harmonious than an equal number of knitting yarn, might have wondered why such colors chosen at random. an elaborate technological superstructure was Because they stress symmetry and needed. mathematical regularity, Ostwald's formulas In the Ostwald and Munsell systems, as in could be easily translated into a computer later adaptations, color is not recognized as program. Sets including every third, sixth, or contextual. Proportion is ignored. We are twelfth of the twenty-four colors in the hue encouraged to conclude that a properly selected circle are identified as harmonious. Indeed, set of harmonious colors, without regard to the "neatly spaced intervals anywhere on the color proportion of each color used, is pleasing for solid lead to harmony" (Ostwald [ 1916] 1969, any purpose at any time. Harmony is implicitly 66). When colors are spaced equidistantly, the an absolute, as if color usage occurred in a ratio of their distances is unity, the ratio Plato vacuum divorced from the affairs of the world. held to be perfect or beautiful. Some of Birren, studying Ostwald's formulas, found, for Ostwald's formulas, such as the ring star, are instance, that "a pale green looks better with a complex, although similarly regular. The deep navy blue than a pale blue looks with a Ostwald formulas cannot be applied to the deep green or olive. Pale yellow looks better Munsell solid because Munsell arranges the with maroon than pink looks with a deep citron" colors differently. (Birren 1969, 46). These dicta are not arbitrary, Color schemes for industry were devised by because, "a number of studies in the field of the Ostwald and Munsell systems, although psychology have verified the observation of interest declined after the 1940s. Jacobson Chevreul that colors look best (a) when they are listed numerous examples, including fabrics closely related or analogous, or (b) when they woven by Herbert G. Antin (Jacobson 1948, are complementary or in strong contrast" (Birren 198). Projects recommended for art students 1969, 35). include painting Ping-Pong balls in 1 mistrust grand generalizations, and so harmonious Ostwald colors and pinning the should the unidentified psychologists who did balls together to look like molecules. That the studies. We cannot determine, once and for activity too has passed into obsolescence. all, which color combinations "look best." An The criticisms that should be offered are entirely yellow-and-maroon world would be obvious. By now, many have been made. Arn- oppressive, because any combination of colors heim, doubting that color harmonies used by can be overused. Cannons of absolute painters can be shown to follow simplistic harmoniousness cannot be reconciled with rules, wisely pointed out that any given set of human propensity for growing tired of anything colors can be used to create either a good at some point. People like variety, at least in painting or a bad one (Arnheim 1956, 337). our society. Stuart Davis, who regarded the defining Munsell's and Ostwald's theories of color characteristic of modern art as its manner of harmony also have a serious technical defect 276 Systematizers and Systems

not remedied in improvements to the systems. Color sets busy clashing in the 1940s (most The mathematical ratios held to define sets of no longer clash) included blue and green, pink colors as harmonious do not correlate with and orange, pink and purple, orange and pur- anything visual. In the Ostwald system (Mun- ple, and red and green. Why do some colors sell's is similar), a set of colors is identified as clash? We need not look for some oddity of harmonious if, and only if, its members are eye or brain that might cause certain color spaced in one of several prescribed manners. combinations to look inherently unpleasant. The determining factor is location in Ostwald's The coinage tells about the beliefs of the color solid. The harmonious sets, as a result, observer, not about the phenomenological show no perceptible family resemblances. world. Goethe pronounced the combination of Assume that a thousand harmonious and a blue and green "repulsive" (Matthaei 1971, thousand nonharmonious color combinations 260). He never bothered to explain why we are created from the Ostwald color solid. A are not offended by the blue sky and green viewer will be unable to sort one group of sets grass. from the other by eye, although in theory har- Tastes change. Clash was thought to monious combinations ought to look better. identify a significant aspect of the Because location, rather than color, is cru- offensiveness of some color combinations. The cial, and no two theorists arrange the colors word provided a way to convey the message within solids in the same manner, a harmoni- that certain colors were not to be used together. ous set of colors according to the Ostwald sys- It gave viewers the comforting feeling that tem is not a harmonious set in the Munsell objective reasons existed for their taste. The system. Nonvisual criteria are presented for arbitrary nature of the concept was veiled so assessing visual phenomena, an eccentric aes- successfully that neurophysiological and thetic principle. The educated eye is a psychological explanations of color clash are meaningful concept. We can learn from Japa- still sometimes sought. nese art, Islamic art, and African and Amerind Baudelaire identified good taste as textiles that individuals and whole societies conventional taste, intolerant of deviation from have been unusually refined in their use of its canons. What was liked or disliked about color. We cannot assume that all creative color in the past is less interesting than the colorists use the same combinations of colors or vehemence with which taste is defended. Red, kinds of combinations. said to enrage bulls, was for several centuries provocative to human beings. The color was regarded as offensive if used in excess, in Color Clash excessively bright variations, or in the wrong Credible evidence is difficult to find about places. Goethe was certain "it may be safely which color combinations look best. A well- assumed, that a carpet of a perfectly pure deep defined body of popular belief purports to blue-red would be intolerable" (Matthaei 1971, identify which look worst. Combinations at 172). Ostwald objected to large areas of red in the bottom of the hierarchy for harmony are Pompeiian wall paintings although today, as said to clash. Clashing, like harmoniousness, Arnheim pointed out, we find Matisse's Red is a musical analogy. With lurid intimations of Studio unobjectionable (Arnheim 1956, 338). clanging cymbals and cacophony, it implies Taboos about red are often disguised socie- that some color combinations are comparable tal taboos about sensuality. Fashion once to musical discords, or that certain colors are decreed, for example, that mature married wrong in juxtaposition to one another. women might wear bright red dresses but Systematizers and Systems 277 young girls should not. We correlate black and tems of private symbolism devised by patients white with wrong and right, including who do not recognize or will not say what wrongness and rightness in sexual behavior. they really think. This is a comment on the The association of red with passion is more nature of symbolism, not just mental health. indirect. We link red with blood, because blood Societal systems of symbolism, like private is red. Blood was once regarded as the seat of versions, say things in roundabout ways. the passions, one of the four cardinal humors Traditional color symbolism provides an thought to explain human disposition. Human illusion of a link to a past that does not exist. It behavior can be described as hotblooded or helps us make sense of the world by using the cold-blooded, and the Passion of Christ is tools earlier peoples used to make sense of it. linked with his blood. Men are called "Red means passion," a student tells me, proud red-blooded, or brave and virile. Women are to be recycling an old saw current for hundreds rarely called red-blooded. of years. Other vagaries of taste can be similarly I wonder if anything that distracts attention traced to symbolic associations with colors. from the present is wholesome. The secret of These associations surface even in works on the great colorists may be an ability to look at colors occult. In a 1924 course of instruction of how without preconceptions, without concern for "to become acquainted with the several astral what was thought about color in the past. The colors perceived by psychic vision," C. social aspect of color symbolism is invidious. Alexander contended that the adept can see Can racial harmony be achieved in a society that astral bodies around human beings. The colors equates good and bad with white and black, also of these auras reveal the conditions of souls uses white and black to code people, and has too through familiar symbolic associations: "Astral many literary ideas about the special status and White is the antithesis and absolute opposite of purity of white? Astral Black . . . . Love gives out the astral If people are to be sorted by skin colors, color of Crimson . . . . The Astral Color of Blue the actual colors should be used. Citizens can represents thoughts and feelings of the class fill in questionnaires by identifying themselves generally known as `spiritual' or `religious' " as members of the somewhat pinkish medium (Alexander 1924, 33-40). beige race, the darker burnt sienna with a yel- Symbolic associations with colors differ in lowish cast race, the pale ocher race, or what- different societies and thus are unlikely to have ever color name best describes the exact skin an instinctual basis. As psychoanalysts and color of that individual. This would give us literary commentators have taught us, anything many more races. Perhaps no two people can be a symbol of anything, requiring only a would belong to the same race. The situation rationalization sufficient to link symbol with would be humorous, ridiculous. But we need the symbolized object. Black is the color of reminder to think carefully about what we take mourning in our society; white has been used in seriously. China. The usage can be explained in each case. For Bitten, "Munsell's work is in every way I see no barrier to a society in which other classical. It has developed permanent values in colors served the purpose. Red might be the the art and science of color" (Munsell 1969, color of mourning, on the basis that shedding 6). For Florence Elizabeth Wallace, Munsell's blood can be equated with death; blue because Color Notation was "an unscientific, meta- the dead go to a heaven located in the sky; physical, but interesting little book" (Wallace green because grass grows on graves. 1927, 1). Munsell was not a scientist. He did Neuroses, as Freud described them, are sys- something original by assembling and grading 278 Systematizers and Systems

a large collection of colors, fascinating to look of the argument lies in its circularity. Those at. Ostwald was a physical chemist whose ideas who do not conform can be dismissed as not about color were often based on what was really great artists. The theorist is certain to be scientifically obsolete or scientifically right although the artist may be wrong. questionable. Jacobson devised a system for keying the The theories of Munsell and Ostwald, hailed general type of color used in a painting. Among by advocates as "doctrines" (Ostwald [ 1916] fifty works listed under "key" categories, 1969, 79), survive today as a foundation for Renoir's Madame Charpentier and Her much of what is taught about color in art Children, El Greco's View of Toledo, and schools. They also survive as a mystique. Botticelli's Three Miracles of Saint Zenobius Bitten, an impassioned interpreter of the (all three in the Metropolitan Museum of Art), mysteries, rarely failed to include respectful are said to be in the same color key. Matisse's capitalization when mentioning "The System" The Blue Window (New York, Museum of or the "Principles of Albert H. Munsell" Modern Art), van Gogh's La Berceuse (Boston, (Munsell 1969, 70-71). Museum of Fine Art), Gauguin's Tahitian The rambling aesthetic of either system is Woman (Chicago Art Institute) and Giotto's The vague and ambiguous. Birren found it was Epiphany (New York, Metropolitan Museum of "Munsell's contention that order assures Art) are similarly placed in a single key. beauty" (Munsell 1969, 69). Ostwald added The rankings are disheartening to anyone that we can "establish this basic law: Harmony familiar with the paintings. The differences in = Order" (Ostwald [ 1916] 1969, 65). Wearing color usage between Madame Charpentier and his own hat, Birren expanded these insights into Her Children and View of Toledo are more a series of incantations to be taken to heart worth reporting than whatever similarities rather than subjected to scrutiny: "There is Jacobson believed can be found. A theory harmony in white, grey, black . . . . There is compelling a conclusion that Renoir and El harmony in pure color, shade, black . . . . There Greco use color in the same way, that Matisse's is harmony in tint, tone, black . . . . There is use can be equated with Giotto's and Vermeer's, harmony in shade, tone, white. . . " (Birren or that Gauguin's colors are similar to van 1969, 54-62). Gogh's is worth rejecting on that account alone. "The Principles" too often lead to dis- We know a great deal about the color theories couraging results. Jacobson, following the Ost- of Renoir, Gauguin, and van Gogh. They were wald system, analyzed the use of color in not using the Ostwald system. twelve well-known paintings .4 He concluded Jacobson wanted to prove that all color in that "the great painters have long used, and all art is the same, unvaryingly a type of color that Ostwald has at last accurately described, Ostwald would have approved. But all color in the fundamental principles of harmony" all art is not the same. Artists have had a wide (Jacobson 1948, 157). This is the unconscious variety of ideas about color. Harmony in Blue, advocate argument, a variation on argument the Matisse painting analyzed by Jacobson, by appeal to authority. Favored by the fol- was seen in a color reproduction. Commercial lowers of Freud, the argument holds that a the- color reproductions of paintings, like color ory can be validated by showing its relevance to photographs, rarely match the colors in the great artists of the past or to their work. The original work of art. Matisse, in any case, said artists may have been unaware of what they he did not use a color system. were doing and may have expressly said they Problems of a more practical order arise were doing something else. The staying power from the limited number of samples available. Systematizers and Systems 279

Jacobson found in El Greco's View of Toledo in cities today. "some greens that are several shades darker" For the classroom, Birren advised two types than any in Ostwald's manual (Jacobson 1948, of color, each calculated to tranquilize, because 162). Richard F. Brown noted that the colors in "it is one thing to make children happy and Pissarro's paintings seldom match the chips in quite another to serve the best interests of child the Ostwald manual "because of the vision and child welfare." For elementary unavoidably limited number of samples in the schools "pale yellow, pink, and peach" were manual," although it is nonetheless true that "a advised. For high schools cool green, blue, or judgment made by means of a carefully gray set a more intellectual tone. "Ivory and specified and standardized color system is pale yellow have been found excellent for certainly more reliable than one made corridors," but in the classroom, "the two best independently by an eye subject to all the hues have been found to be pale blue-green and deceptions of surroundings, colors, changing peach." Reds must be avoided because "red is illumination, etc." (Brown 1950, 12). inciting to activity." Birren's preferred choices For Birren, " `natural laws' of color harmony elsewhere became known as institutional ivory, exist without question," a compelling argument institutional green, institutional peach, for systematizing the use of color (Birren 1961, institutional gray. 39). In "Functional Color in the Schoolroom" These ideas, for a period, were widely (1949), Birren outlined what he described at influential. In 1948, "Faber Birren developed that time as the science of functional color. The manuals of standard color practice for the shore application of the purported science was to the establishments, surface vessels, and submarines designing of classrooms, because "a sober of the U.S. Navy. This was followed in 1952 by study of school children and school a similar report for the Coast Guard, and in environments will stress the need for an 1966 by a report for the U.S. Army." In 1955 objective approach to color rather than a the U.S. State Department "sent him as an subjective one" (Birren 1949, 136-38). Color expert on functional color to an International functionalism has a well-defined set of rules, Congress in Rome on work productivity and and although "a relatively new science, its safety" (Birren 1969, 8). Governments by their progress has been rapid, for the benefits of its nature are the keepers of morality, conservative application may be definitely proven through in color theory as in much else. research studies and clinical tests. While The science of functional color faded when individual accomplishments in the fine arts are Birren incorporated his ideas into a movement often difficult to evaluate, functional color he called Perceptionism, "an advanced art of stands or falls on its measurable results." color" (Birren 1961, 63), and "a joy to study The message of color functionalism in 1949 because it deals with human reactions" (Birren was to avoid high chromaticity, because it is 1961, 97). Birren exhibited Perceptionist "difficult to read a book against the competition paintings at the National Arts Club in New York of bright colors." Bright colors also lead to in 1948, and the movement presented its own eyestrain, and as the child twists to avoid this, rules, laws, principles, and copying exercises. the twisting "may affect posture and have an Those who knew were always Perceptionists at adverse effect upon the growth of young heart, and "many of the great artists of history bones." Furthermore, technical advances in have understood Perceptionism and The Law of modern illumination render "the glare of white Field Size or Proportion-perhaps not in a walls intolerable," although white is one of the scientific way but through an innate genius" most popular colors used for apartment walls (Birren 1961, 69.) 280 Systematizers and Systems

For beginning students, Birren's Creative In art schools and college art departments, Color includes color charts and diagrams to be the pedagogical posture that confounded imitated, although, "frankly, these preliminary neatness, following instructions, and careful exercises bear resemblance to the playing of copying with discipline survived longer in the scales in music and to the parsing of sentences teaching of color than in the teaching of in grammar" (Birren 1961, 9). Trickier copy drawing or painting. It was not limited to those work for the advanced student includes practitioners on the fringes. During the 1920s reproductions of thirty Faber Birren paintings coloring diagrams was an integral part of the that "are meant to speak for themselves, as color class taught by Johannes Itten at the original and striking examples which point to Bauhaus. Exact imitation was the pedagogical new directions" (Birren 1961, 10). Anyone goal. In one exercise, "all three basic colors "may copy them if he wishes," although Birren must be clearly represented, checked by the reminds us that "from experience I doubt if even teacher, and painted into the three parts of an a skilled artist could take a look at these equilateral triangle" (Itten 1963, 41). In addition illustrations and then, on his own, do as well" to color triangles, color stars and color circles (Birren 1961, 96). An appendix, entitled "The (Itten's had twelve sectors) were used. Faber Birren Palettes," assists the aspiring The pedagogical model tells us to imitate copyist by providing listings of the colors used our betters rather than trusting our ability to in each painting. learn by observing. Few other formats for Elsewhere, Birren displayed a reproduction teaching about color have been fully explored, of a Jackson Pollock painting, an example of and genuine inventiveness, I fear, is limited to the bad ends awaiting those "without dis- the more or less mad. Brooding over the colors cipline" (Munsell 1969, 43). Printed opposite of his scraps of knitting yarn, van Gogh must the offending Pollock is a work by Andrew have learned something. Another eccentric, Dr. Wyeth, who exemplifies "conservatism, tradi- Barnes, became a collector of the work of tion and discipline in American art and color Matisse after becoming wealthy from the expression." For Birren "a musician needs to invention of Argyrol, a vile-tasting swab used know a great deal about the elements of music on children's throats. The listings Dr. Barnes before he can compose. Then he can forget the compiled of the colors in Matisse paintings may rules. The same with color harmony and the appear pointless to scholars (Barnes and de principles of Albert H. Munsell" (Munsell Mazia 1933). They show, however, that Dr. 1969, 70). Problems begin when those without Barnes looked carefully at the colors. the imprimatur of genius break the rules. CHAPTER 31 The Logarithmic Gray Scale

In the case of the Ostwald Color Solid, we have visually adjusted intervals according to Fechner's Law of Sensation which states briefly that in order to change visual perception arithmetically the stimulus intensity must be changed geometrically. Charles N. Smith, Student Handbook of Color

gray scale is a chart of color swatches. usually enough. This abbreviated gray scale It has black at one end, white at can be abstracted from a finer scale by select- the other, and a series of graded grays ing, say, every tenth, twentieth, or hundredth in between. A scale can be constructed by step. collecting many samples of achromatic gray Wilhelm Ostwald proposed a more ornate and ranking them from lightest to darkest. In methodology for constructing gray scales, the finest possible scale, the difference in value based on the Weber-Fechner law. Intervals (degree of lightness or darkness) between one between grays (measured in terms of their gray and the next is so slight as to be nearly reflectances of light) are logarithmic. The imperceptible. From sufficient distance (a small reasoning that led to this widely circulated scale distance in this case), the scale looks like an illustrates in miniature the manner in which unbroken continuum blending from white unnecessarily convoluted reasoning about color through gray into black. A scale with many overwhelms simpler, more visual, small gradations approaches this continuum understandings. more closely than another with fewer and larger Gustav Theodor Fechner (1801-87), who steps. inspired Ostwald's marriage of color with log- We rarely need a gray scale with thousands arithm, was trained in biology and became of small steps; eight or ten larger steps are professor of physics at the University of Leip-

281 282 The Logarithmic Gray Scale

zig. After 1843, Fechner, whose major work that human beings find the proportions of the was Elemente der Psychophysik (Elements of golden section more beautiful than any other Psychophysics), turned to the study of proportions. Fechner had discovered this philosophy, experimental aesthetics, and psy- aesthetic law by asking subjects which of ten chophysics. He hoped to demonstrate that rectangles of different proportions they liked aesthetics and the arts were governed by scien- best. Bosanquet, like Croce, was unpersuaded, tific laws. arguing that the actual tendency was "to prefer Fechner's philosophy was influenced by a form that was not extreme in a given series," Friedrich von Schelling (1775-1854) and and that "most of the persons began by saying Johann Herbart (1776-1841). He believed that that it all depended on the application to be animate and inanimate objects have souls made of the figure, and on being told to comparable to those of human beings. All souls disregard this, showed much hesitation in are animated by a greater world soul similar choosing" (Bosanquet [ 1892 ] 1957, 382). (although Fechner did not use this comparison) In 1839, in a disastrous experiment to the Atman of Hindu philosophy. Natural law designed to uncover the nature of colored is the revelation of God's perfection, therefore , Fechner stared too long at the sun. perfect itself, a dubious though once-popular He lost his sight for three years and, on syllogism, dubious in that this perfect God recovering it, claimed to be able to see the souls created imperfect men and women and may of plants. The experience led to Fechner's have also made an imperfect universe governed turning from physics to psychophysics, and to by imperfect laws. publication of Nanna, or the Soul Life of Plants The German philosopher Arthur (1848). Schopenhauer (1788-1860) had defended Fechner attracted a heterogeneous following, Goethe's theory of color. Schopenhauer which included the philosopher-psychologist understood Goethe to mean that colors, as William James. P. D. Ouspensky, GurdjiefPs Aristotle concluded, are composed of mixtures major follower in the United States, described of the opposites of darkness and light his own views as essentially similar to (Bosanquet [ 1892 ] 1957, 366). Herbart, Fechner's (Ouspensky 1947, 188). Ouspensky closely associated with Schopenhauer, added was attracted by the transcendental cast of the call for "a science of color harmony, like Fechner's writings. Psychologists drawn to that of harmony in music" (Bosanquet [ 1892 ] Fechner were more interested, as was Ostwald, 1957, 370). The science was to determine in schema that promised methods for measuring which color combinations were beautiful and sensation. If sensation could be measured, the provide scientific proof of its findings. door could be opened to a mathematical Unlike Schopenhauer and Herbart, Fechner understanding of subjective experience, of the experimented, though his experiments and human condition. purposes were often not greeted warmly. The Proposing an equation for the relationship Italian philosopher-statesman Benedetto Croce between stimulus and sensation, Fechner said (1866-1952) complained about Fechner's that the intensity of sensation is proportional to assiduous discovery of too many so-called laws a multiple of the logarithm of the stimulus. This that too neatly supported his scientific aes- equation might apply to, say, stimulation of the thetic. Croce wondered why Fechner eye by light waves that cause a sensation of experimented at all if he knew in advance what seeing red: he wanted to find (Croce [ 1909] 1968, 394). An oft-cited experiment persuaded Fechner sensation intensity = C log stimulus intensity The Logarithmic Gray Scale 283

C is a constant to be experimentally determined a logarithmic scale, a convention with no for each sense modality. Because the German known bearing on vision (Blaker 1969, 1). anatomist and physiologist Ernst Heinrich Beyond this, the hypothetical experiment Weber (1795-1878) had proposed a similar includes no provision for sorting out an ill- constancy in 1834, the formulation is called the considered hodgepodge of factors. As the light Weber-Fechner Law. It has been criticized on grows progressively brighter, the viewer will logical grounds, and experimental evidence grow progressively more tired from the unnat- fails to support Fechner's claim that this is a law ural task of sitting in the dark looking at it. If of universal applicability. progressively greater stimulation is required to Weber experimented primarily with touch produce a similar brightness change, is eye and to a lesser extent with sound. The fatigue a contributing factor? A determination WeberFechner Law became popular among might be made by running the experiment researchers in vision and hearing. Through backward, stepping down a bright light until Wilhelm Ostwald, it was introduced to color it is no longer seen. If eye fatigue is not sig- theorists, with dubious results in the nificant, the decreasing steps (when the construction of gray scales. An example often experiment runs backward) should match the used to illustrate the application of Fechner's increasing steps (when the light changes from law to visual experience asks that we imagine a dim to bright). darkened environment in which a faint light A confrontation between a human being and gradually grows brighter.' a gradually intensifying light source involves When the light reaches an intensity strong several visual thresholds. Fechner recognized enough to be noticeable, it is said to have just one, perhaps because the proposed reached the threshold (of vision). The threshold experiment cannot be carried out from full varies among individuals and may vary for any darkness to full light-ethical and safety factors individual at different times. At a further intervene. Yet the entire gamut is needed to increase of intensity, the light becomes understand the scope of the phenomenon. observably brighter, by the smallest amount of Beyond the level of tolerable intensity, light noticeable difference. A third increase of causes, progressively, eyestrain, pain, temporary specific intensity again causes the viewer to blindness, permanent blindness, and death, each report awareness of an increase in brightness. a threshold of another order. Human beings By continuing sequential increases, a scale tolerate radiant energy only within can be constructed. Between any two of its circumscribed limits, and a series of changes of steps, a just noticeable difference in brightness state can be anticipated in the eye viewing the of that light exists for that observer. Fechner light. held that the increases in intensity required to Similar limits can be seen in exposing produce the steps grow progressively larger camera film. At a point depending on the speed and show a logarithmic progression. When (light sensitivity) of the film, maximum light is faint, a tiny increase makes a overexposure is reached. The film, depending difference. As the light becomes brighter, on whether negative or reversal, produces an progressively greater amplifications are image either all black or all white. Further required to produce a noticeable difference. increases of light intensity create no further Fechner was a physicist, familiar with the change in the image recorded, unless the use of logarithmic units on scales for measur- increase is so great that it generates enough heat ing radiant energy. In texts on optics, the to destroy the film. entire electromagnetic scale is often shown as The eye, damaged by too much light, is also 284 The Logarithmic Gray Scale

unable to tolerate the coherent light of a laser Other thresholds for eyestrain, although not beam. Despite human passion for order, we are germane to Fechner's experiment, are fortunate that atoms under ordinary encountered every day. Common causes circumstances do not act in lockstep unison in include looking at fine detail, looking at lights emitting radiant energy. The pure beam of the that blink (rapid changes from high to low laser results when stimulated emission is used intensity), or an environment combining to cause them to do so. several intensities of illumination. Eyestrain At the lower reaches of Fechner's experi- from reading with insufficient light probably ment, one of the first visual thresholds to make comes from stressing the eye in scotopic mode itself manifest is between scotopic and pho- by presenting finer detail than it can topic modes. The eye sees in the dark by comfortably accommodate. means of the rods near the perimeter of the Each of these visual thresholds marks a retina (scotopic vision). In a lighted environ- transition from one state to another when a light ment it sees largely with the cones surround- is stepped up or down. None of them can be ing the blind spot at the retina's center equated with the single threshold Fechner labels (photopic vision). The point at which the the threshold of vision, a threshold that does not environment is experienced as lighted, signal- exist. Fechner's concept is unfortunately ing the transition from rod to cone vision, varies framed, because the point at which a faint light among individuals. In the dark the cen- is first seen in the dark is only a threshold for tral portion of the retina is blind because its seeing that light. No absolute threshold of cones are inactive. vision exists for the functioning eye. At no point Whether or not an evolutionary relic, sco- do we stop seeing. topic vision is crude, allowing limited percep- Feeling no pain is possible, as is hearing no tion of fine detail. It does not allow a dim star sound. Genuine thresholds exist for those or other faint light to be seen in the dark if the sensory experiences. Silence has no visual viewer looks directly at it. The viewer must analogue, and no threshold exists for vision or look, instead, about twenty degrees to the side color. During periods of consciousness, the of the light, a phenomenon known as averted functioning eye unceasingly engages in seeing, vision. Unlike the bright/dim of the automo- even if watching just the nominal black of the bile headlight, the eye's two systems for see- night. An extension of the brain, the eye never ing differ in the way that each functions. One turns off, and REMs (rapid eye movements) question is whether a just noticeable difference have been shown to occur during sleep. in photopic vision is registered as just notice- Differences between pigments and lights able in scotopic mode. Because of the greater were pointed out by Goethe in his criticism crudity of scotopic vision, I suspect it is not. of Newton and later were much discussed by Visual thresholds beyond those for scotopic late-nineteenth-century painters. Looking at and photopic vision include those for eyestrain lights in the dark, as in Fechner's proposed and pain, both of which point to the limits of experiment, is not comparable to looking at visual acuity. Within the hypothetical colored pigments or colored objects in day- experiment, eyestrain would occur at two light. The brightest white pigments are less points. The first is when the amount of light is intense than the brightest white lights, which less than optimal. The second is when the light bears on the representation of light effects in is so bright that the viewer experiences it as art. The sun can never look as bright in a paint- glare. ing as in actuality. For Ralph M. Evans, the dis- The Logarithmic Gray Scale 285 tinction between pigments and lights Proceeding from these givens, it migh seem invalidates Ostwald's adaptation of the that the gray midway between the blacl and the WeberFechner law, for in Ostwald's "practical white ought to have a reflective indei. of 50.5 reduction of [Fechner's) theoretical percent, an average between thf extremes. considerations to a set of material paper Ostwald argues that it will be univer sally standards it was necessary to use actual agreed 50.5 percent looks too light. The proper pigment colors, and they departed so widely gray, "halfway" between white anc black, from the theory as to represent a different shows a reflective index of 10 percent system" (Evans 1948, 216). (Jacobson 1948, 195; Ostwald [ 1916) 1969 Fewer thresholds must be taken into 25). Thus, "if one were to arrange [the step; of account if the Fechnerian experiment can be the gray scale] according to equal (arith metic) adapted so that colors are used instead of steps of contents of white, one woulc obtain lights. Unfriendly critics of the early twentieth too many steps at the white end and too few at century insisted that the colors in French the black end" (Ostwald [ 1916] 1969 29). The Impressionist paintings were too bright and reflective index of any gray b, inter mediate might injure the eyes of viewers. But material between two others (a, c) can be found by the objects, including spots of pigment, are less formula alb = b/c. than perfect reflectors. Their colors rarely If gray a has a reflective index of 1 percent injure eyes. Strong light of any color has that (black), gray c a reflective index of 100 percent potential, whether direct sunlight (as Fechner (white), then intermediate gray b has a painfully learned), when rays are brought into reflective index of 10 percent (1/10 =10/100). focus through a lens to start a fire, or in the This is a geometrical progression, and Ostwald coherent beam of a laser. If, in the case of found that "where brightness differences form a light, each threshold represents a shift of gear geometrical progression, only then do we in the visual apparatus, seeing colors in experience corresponding grays as being daylight has only one gear. visually equidistant" (Ostwald [ 1916 ] 1969; An abridged English translation of Ostwald's 26). Ostwald found that the total "number o1 Die Farbenfibel is edited by Faber Birren, and distinguishable steps of gray under normal the system is also outlined in Jacobson's Basic conditions amounts to several hundred," though Color (Ostwald [ 1916] 1969; Jacobson 1948). other writers have given numbers ranging from In adapting Fechner's theories to the five or ten to seven hundred. construction of a gray scale, Ostwald began Whether Ostwald conducted tests or just with a white surface and a black surface. The made his own guesses about what most people white is assigned a reflective index of 100 would guess, determining an average is not percent. This is a nominal index. Ostwald accomplished by taking a vote. The midpoint explains elsewhere that no white surface between indices of 1 and 100 is 50.5, a matter reflects 100 percent of the light falling on it. of computation. If inconsistent with what most Ostwald's black surface absorbs nearly all the people estimate, we need to know more about light falling on it and is assigned a reflective most people's competence at estimating index of 1 percent. Why the less-than-ideal reflective indices. black receives an index of 1 percent, rather than Lightness and darkness are inherently im- a zero, while the less-than-perfect white precise terms, and half as dark has no firmer receives an index of 100 percent, not a lower meaning. An observer shown black, white, and number, is not clear. a single gray has no basis for answering such 286 The Logarithmic Gray Scale

ill-conceived questions as whether the gray is though its logic is rarely understood. Ostwald "halfway" between the black and the white. The began with a scale of twenty terms, a geometric question can only mean halfway on a properly progression (figure 31-1). The scale has ten constructed gray scale. All steps of the scale descending steps between 100 and 10; another ought to be shown. Although an untrained ten between 10 and 1. Each term in the series observer may not be sensitive to the issue, (Ostwald's numbers have been rounded off) can judgments about colors are made most be arrived at by multiplying the previous term effectively if all colors in question are seen. If by 0.79. an observer moves progressively farther away The ratio of any three consecutive terms is from the scale, a point will be found at which a/b = b/c, a ratio that also holds for any three any two given successive steps no longer look terms equidistantly spaced. For the first, third, sufficiently different to allow them to be and fifth terms: 100/63 = 63/40. For the second, identified as two different grays. If this critical sixth, and tenth: 79/32 = 32/12.6. viewing distance is the same for all steps, the Ostwald's next step, for which I see no steps are properly balanced and the middle step purpose (it shuffles numbers without changing is correct. relationships), was to take the mean between More exactly, the steps of the gray scale are all terms in the series (Ostwald [1916] 1969; properly balanced for a particular viewer. We 26; Jacobson 1948). This leads to twenty-five do not know whether a universal gray scale is norms of gray, keyed to all letters of the possible: a scale balanced so that, for any alphabet except j (figure 31-2). Because the viewer, a single point can be found at which geometric progression contains twenty terms, discriminating between successive steps is no only nineteen means are available between longer possible. For a given individual, is ability successive terms. The six additional means (uz) to discriminate among dark colors equal to were apparently arrived at by computation. ability to discriminate among light colors? Ostwald later deleted them without explaining People who suffer from night blindness have a why they were added. They inconsistently reduced ability to see in semidarkness. They imply blacks of less than 1 percent reflectance. might have greater difficulty distinguishing fine Ostwald had said the scale could include no differences among dark grays than among light black surface that reflected less than 1 percent grays. of the light failing on it. Ostwald used several steps of computation Ostwald's next step was to abbreviate the to arrive at the final indices for his logarithmic norms of gray by removing every alternate gray scale, a scale widely circulated today term (Ostwald [1916] 1969, 27). This resulted The Logarithmic Gray Scale 287

in a construction he called the practical gray black (1 percent reflectance). The ability of, scale (figure 31-3). The explanation for say, gray a to reflect 36 percent of the light abridging the norms of gray was that a scale of falling on it accounts for its appearing to be that twenty-five steps is too long to be practical. Its particular shade of gray. Whether Ostwald is steps are unnecessarily narrow, and thirteen using an absolute or a normative standard for grays will suffice (Ostwald [1916] 1969, reflectance is often unclear. If absolute 23-27). measurement is implied, a particular gray is To make the practical gray scale more illuminated by a known quantity of light. The practical, Ostwald discarded 38 percent of it: amount of light reflected is measured, and a the last five of the scale's thirteen steps were ratio calculated between input and output. By deleted, dispatched to some limbo from which this standard, a surface with a reflective index they will not emerge. The result is a scale of 100 (100 percent reflectance) reflects 100 ending with p, said to be equivalent to the percent of the light shining on it. reflective index of black printer's ink (3.6 More likely, Ostwald's reflective indices are percent), because "a deeper black than p could normative or nominal, a detail easy to over- not be created by normal printing processes" look in the welter of confusing computations. (Ostwald [1916] 1969, 27). Ostwald's final Ostwald pointed out that the closest approxi- eightstep gray scale has reflective indices as mation to ideal white is a dull coat of barium shown in figure 31-4. Values for grays a to p, sulfate, the surface taken as a norm 100 (Ost- marked off on a ruler or other arithmetic scale, wald [1916] 1969, 24). The actual reflective will not fall equidistantly. On a logarithmic index of this surface, which is unspecified, may scale they fall equidistantly. For Ostwald, "a be 80 percent, 96 percent, or some other thorough examination of this gray scale shows number. White of 100 percent reflectance, by that the distances are indeed experienced as this normative standard, matches the reflective being equally spaced to the eye. Only the lower capacity of barium sulfate. A gray of 36 percent steps give a more crowded impression, in view reflective index is measured by deviation from of the circumstances that the law of geometric barium sulfate, not in absolute terms. progression no longer exactly represents the Why is this unnecessary complication facts among dark colors approaching black" included? If the actual reflective index of (Ostwald [1916] 1969, 27). barium sulfate is, say, 85 percent, this number could have been used. Adjusting at this point would require recomputing all of Ostwald's reflective indices, to what entire effect is unclear. The logic of the scale-its mathematical logic-is a greater problem. No reproduction will be found in print showing the grays that correspond to Ostwald's entire scale of In the completed scale, the reflective index logarithmical reflective indices. The sector of assigned to each gray locates it within the the logarithmical gray scale that is shown range from white ( 100 percent reflectance) to usually runs from atop. The entire scale can- 288 The Logarithmic Gray Scale

not be reproduced, because it has no lower between 1895 and 1900. In the laboratory a end. The numerical progression on which the blackbody (which need not be black) consists of scale is based continually diminishes at its a hollow sphere or a tube with closed ends. It lower (black) end, but never reaches zero (see has a small hole in its side. The assumption is figure 31-1). The equation a/b = b/c cannot be that radiation entering the hole will bounce solved if c is zero and a and b must be greater about on the curved interior walls until than zero. absorbed, with little or no probability of If the limit is appropriate for the bouncing out through the hole again. The hole circumstances, scales can be one-ended. The is a perfect absorber (or transmitter) of light. Kelvin scale, measuring temperature, has a During the nineteenth century, the lower end at absolute zero (0°K) but no upper distribution of radiation from a heated end. We do not know what the highest blackbody was found to be inconsistent with the temperature would be or if a highest possible laws of classical physics. To explain this, temperature exists. The numerical progression Planck proposed what later became known as used for Ostwald's logarithmical gray scale, the quantum and laid the foundation for running from 100 to 0 but never reaching 0, quantum theory. The blackbody also serves as a makes sense if Ostwald believed that some model for the black holes cosmologists believe natural substances have a reflective index of 100 may exist in outer space, regions of gravitation percent (absolute white), but none have a so intense that even light is unable to escape reflective index of 0 percent (absolute black). (Hawking 1988). He believed the opposite. He never saw the Because of interest in the blackbody in inconsistency of his computations, or he said physics, we need not look further for the nothing in his zeal to prove that color could be source of Ostwald's assumption that complete reduced to simple mathematical symmetries. absorbers of light can be created. Why a mirror Barium sulfate is assigned an index of 100 was not suggested as a perfect (or the most based on Ostwald's argument that this is the nearly perfect) reflector of light is unclear. most reflective surface known, and absolute Whatever the case, Ostwald's scale of white does not exist. No surface reflects all reflective indices would be more compatible light falling on it. Ostwald is more confident with his beliefs about absolute white and about absolute black. He describes a method for absolute black if it were turned upside down to creating it. Under laboratory conditions, "a become a scale of absorptive indices. Absolute black that will not reflect any light can be black would be represented by 100 (it absorbs produced by making a 4-inch cube-shaped box all light). Absolute white (0 percent of dull black painted cardboard, with the black absorption) would have no place on the scale, surface facing inside, then cutting an opening given that the logarithmic progression on about 3/4-inch square in the center of one side" which the scale is based never reaches 0. But (Ostwald [1916] 1969, 24). The viewer looking this is consistent with Ostwald's assertion that into this hole sees an absolute black reflecting there is no white with a reflective index of 100 no light. The problem about this black of 0 percent (0 percent absorption), and its closest percent reflective index is where to put it. approximation is a coat of barium sulfate. Ostwald's logarithmic gray scale approaches, This adjustment would not resolve another but never reaches, zero. inconsistency. A midpoint is a place equidis- Ostwald's instructions for creating an abso- tant between fixed extremes. A scale with only lute black are those for building a blackbody one end has no midpoint. Ostwald undertook in physics, a model of interest to Max Planck to develop a gray scale that would identify the The Logarithmic Gray Scale 289 midpoint between white and black, the gray half words imprecise. The names white and black as light as white and half as dark as black. The have another kind of ambiguity. They tend to be logarithmic gray scale, offered as a superior applied to extremes, often to the lightest and solution, suggests no solution is possible. As darkest colors in an environment. A movie the scale is extended to progressively greater screen that observers in an illuminated room lengths, its predictions grow progressively more agree is white looks less so if compared to the absurd. If long enough to include all of whiteness of a beam of light falling on it. Ostwald's norm grays from a (white) to z Subjective understanding of what constitutes (black), the Ostwald gray scale predicts that the white varies. The color class is relativistic or median point between them is gray n (5.6 arbitrary, a limit difficult to reach beyond. In percent reflectance). But gray n, by Ostwald's theory, absolute white can be defined as the report, is just one step lighter than the black of maximum amount of light the human eye is able the blackest printer's ink. I conclude Ostwald to tolerate without damage. We have no exact amputated the sector of the scale beyond P (3.6 idea of what that amount is. percent) because the scale looks ridiculous if The darkness of night, though the lightest as seen at its entire length. The printing industry, well as the darkest color of that hour, is not he said, uses no ink of less than 3.6 percent called white. We call it black, without reflectance. reference to whether it looks like an absolute black. The elusiveness of white and black has its counterpart in other color ranges. White and Black Identifying a color as black, white, or red, we Ostwald contended that if human vision were forget having seen blacker blacks, whiter sufficiently acute, we would be able to see that whites, redder reds. no two gray spots in the world are exactly alike What is forgotten is that any color name in value (Ostwald [ 1916] 1969, 21). The identifies a range rather than a single tonality. I proposition implies that if a sheet of paper is can imagine ways to program a computer so painted a uniform gray, cutting the paper in that the machine would always be aware of half establishes a difference in darkness among when I meant a single shade of, say, pale blue- previously value-identical halves. It also violet and when I meant a range of thousands implies that only one color spot in the world is, of varieties of that color. Communicating with say, absolute white, and does not address the human beings is more difficult. The English issue that at twilight this spot might look language is not structured to encourage grayish rather than perfectly white. Ostwald speakers to make the distinction and offers few gives insufficient attention to simple visual easy forms for doing so. phenomena. No object, including a color spot, Forgetting that colors are ranges is looks the same color under all kinds of lighting encouraged by language and by the constraints conditions. We cannot make broad of human memory, which is poor for exact generalizations about what color objects, shades of color. Unless two pieces of navy blue including color spots, really are. Because of the fabric are seen at the same time and place, in phenomenon of optical mixture, a dark gray close proximity to one another, we remain and a light gray can be indistinguishable-will uncertain about whether their colors exactly look alike-at a sufficient distance. match. People take wallpaper samples to paint Light (in color) means approaching white. stores because how two colors look together, Dark means approaching black. In neither case and whether they match, cannot be decided do we know how closely, which makes the unless both are seen at the same time. Most of 290 The Logarithmic Gray Scale

us have learned the humble wisdom of trying iest way to test a gray scale is to walk away on clothes and looking in a mirror before from it until this critical distance is found. Opti- deciding whether to buy them. Seeing, we cal mixture takes over; each gray visually know. Making assumptions without looking merges with those adjacent to it. If grays a and leads us astray. b differ by "the same amount" as grays c and Why memory for color is poor is uncertain. d, this can only mean that each approaches a Preparing the mind for attention to the pres- continuum equally closely, with the number ent implies suppressing the past, an aspect of of intermediary grays distinguishable between the existential nature of the human condition. a and b equal to the number between c and d. This alone is insufficient to explain why The distance at which a and b look alike events, feelings, and words impress themselves (become continuous) should match the distance upon us more indelibly than the small nuances at which c and d look alike. In a properly among colors so easily forgotten by both balanced scale, the distance at which adjacent trained and untrained observers. This weak- steps look continuous should be the same for all ness of memory coexists with a high degree steps, for the lighter grays and for the darker of perceptual acuity. If shown the individual grays. colors in a series, most people can efficiently A convenient gray scale for testing is used arrange them in order, whether by hue, value, by photographers and published by Eastman or chroma. Most people are similarly adept at Kodak (Eastman Kodak 1966, 11). Its steps sorting a series of achromatic grays from light- match those of the Munsell gray scale (Munsell est to darkest. Student Chart, Hue/Value/Chroma), suggesting Despite its refinement, the ability to rank by logarithmically determined steps. An advantage hue, value, or chroma has limits. It disappears to the Kodak gray scale is that each rectangle of at the threshold for optical mixture. If two spots gray touches the grays on either side of it, a of color, which can mean two grays, are seen necessary condition for testing balance. Most from sufficient distance they appear to fuse, as other gray scales are printed with a separation in the case of the building of multicolored between the steps. If the Kodak gray scale is bricks that looks like a single color from far viewed from gradually increasing distances, away. differences between the lighter grays remain The logarithmic gray scale has become visible at a greater distance than differences standard, following Ostwald's scale or the between the darker grays. The center gray is too Munsell Company's adaptation of it. Is the dark, a monument to Ostwald's foray into central gray too dark, an imbalance inspired by logarithms. Ostwald's concern that it might be too light? In The imbalance of logarithmically putting the question to the test, let us assume determined steps rarely attracts attention, as that the just noticeable difference between two several ill-considered conventions are followed grays is constant for a given viewer under in printing gray scales. Most have a separation controlled lighting conditions and proportional or band between one step and the next. A gray to the distance between viewer and object. If I scale is easier to evaluate if each step touches can see a just noticeable difference between the steps alongside it, an extension of that basic two grays at a distance of ten inches, this axiom of color matching. Whether two colors difference will not be visible from a hundred match, or the degree to which they do not, is feet away. most accurately estimated if the colors are A critical distance exists at which consecu- placed so that they touch one another. tive steps are perceived as continuous. The eas- Most gray scales are printed on white paper, The Logarithmic Gray Scale 291

another distraction. Any color is affected by the it in many ways. Yet his adaptation of the colors surrounding it, Chevreul's law of Weber-Fechner Law to construct a gray scale simultaneous contrast. In the usual format for was not a happy experiment. The logarithmic gray scales, the color surrounding the scale (the gray scale-by which I mean the entire scaleis white of the paper on which it is printed) is also not successful in predicting the median gray, the color of one extremity. The viewer and makes progressively more absurd involuntarily reads the contrast between each predictions as its darker end is approached. gray and its white environment (no way to Ostwald's excising of 38 percent of the avoid seeing it), although ostensibly focusing steps at the dark end of the scale makes the attention on the degree of difference between defect less noticeable but does not remove it. successive gray steps. Substituting a black What should have been jettisoned was the background is no solution, though any gray insistence that the scale be logarithmic, that it scale looks different if its background is black conform to preconceived ideas that were instead of white. What is needed is no mathematical rather than visual, that meant to background at all, or a background color prove a point about whether scientific different from black, white, or gray, and neither explanations could be provided for visual light nor dark. In the first case, each step can experience. I suspect we shall have our touch the next, and the steps can extend to the scientific explanations at some future date. A edge of the paper. In the second, each step more advanced science will be needed than is should touch the next and the band of presently available. Certainly a more advanced background color surrounding the scale should science will be needed than that offered by be, say, vermilion. Ostwald and Fechner. A second reason gray scales are rarely Today a sophisticated literature explores the looked at closely is that most people sense how nature of scientific proof. Whether natural laws poorly they reproduce in books, where ideas can be found to explain color, beauty, the arts, about color and color value are difficult to or aesthetic experience is not an interesting illustrate. The Ostwald gray scale is reproduced issue, though still occasionally raised by some in The Color Primer and in Jacobson's Basic psychologists. We smile at the excesses of those Color. The scales in the two books do not look late-nineteenth- and earlytwentieth-century alike. In The Color Primer, one pair of thinkers, including Fechner and Ostwald, who illustrations shows the relative whiteness of expected to find a simple scientific law for zinc white and chalk white (Ostwald [1916] everything and to finish the job by next 1969, 24). Although chalk white (80 percent Tuesday. reflectance) is darker than zinc white (92 We still use the logarithmic gray scale and percent), it looks lighter in the illustration. Two have too few answers of our own for the ques- diagrams identified as 80 percent gray do not tions Ostwald addressed. Constructing a look equally dark and are darker than a third visually balanced gray scale is a worthwhile gray the text identifies as 56 percent gray. endeavor for the National Bureau of Standards, (Ostwald [19161 1969, 24, 25, 28). a scale balanced so that each gray merges with Ostwald was a leading scientist in late- the grays adjacent to it at the same viewing dis- nineteenth-century Germany. In 1909 he tance or angular subtense. A greater need is for received the Nobel Prize in chemistry for dis- standardized tests to allow collecting reliable covering how to oxidize ammonia to yield information about visual acuity for color rela- oxides of nitrogen. His color system, though tionships and color value (lightness, darkness) it followed Munsell's closely, was superior to relationships in a normal population. 292 The Logarithmic Gray Scale

Tests for color blindness identify those Similar scales could be established for ability whose vision is so grossly atypical that we to distinguish between colors close in hue, characterize it as defective. These tests were value, and chroma. The index of, say, 10/20 not designed to measure variations in acuity could mean the ability to see, at a distance of for color among those with normal vision. ten feet from the chart, a difference between Tests for fineness of acuity for small color two colors, or two grays, that most people can differences, including small differences distinguish from a twenty-foot viewing between grays, would be easy to construct, distance. administer, and evaluate. They could be created We do not know whether a correlation exists by adapting tests and methodologies already for a given individual between the just- available and would not be difficult to noticeable difference for different colors. standardize. Imagine that a hundred people are in a high Trained observers are often used in percentile for ability to notice small differences colorimetric studies, raising the question of between steps of gray. Whether they will share whether training improves the ability to a similarly high level of acuity for small distinguish fine differences in color. We do not differences among reds, blues, lavenders, and know that it does. The ability may be inherent. each of the other colors is difficult to predict. The issue can be avoided if the ordinary eye Explanations of anomalous color vision are chart is adapted as a test instrument. Letters in based on the assumption that no such different values of gray could be printed on a correlation should be expected. medium gray background. The test of whether A commercial incentive exists for testing for the individual could discriminate between gray nearsightedness and farsightedness. Eyeglasses 1 and gray 2 could be whether he or she was can be fitted to improve acuity-at a charge. No able to read, at various distances, letters printed such incentive is available to test for acuity in in gray 1 on a background of gray 2. Except for color discrimination. We know no way to the difference in the chart, the test is exactly that improve substandard acuity. If information presently used for nearsightedness and about color acuity were collected from a large farsightedness. population, it would prove its usefulness, I Statistical methods already exist for deter- think. The information would provide a clearer mining an average from individual variations. picture of what to expect from the normal eye These methods are applied to tasks as diverse as and would place color blindness in context. It determining average life expectancy or aver- might help prevent blindness by showing how age ability to see the forms of letters on an eye to look for early warning signals. Certainly it chart. Twenty/twenty vision is the ability to would help in establishing standards for an see at twenty feet what a statistically deter- adequate level of color value discrimination mined average person sees at that distance. when driving a car at night. CHAPTER 32 Primary, Secondary, and Tertiary Colors

In the De Coloribus another writer of the Peripatetic school, probably not Aristotle, names three elementary colors, black, white, and yellow and by most fantastic reasoning he derives all other from them. Florence Elizabeth Wallace, Color in Homer and Ancient Art

ne purpose of color theory is to pre- and earlier efforts. Primary colors are simplex dict the results of color mixture. or modular units that enjoy the status of an Ora nge, for example, is predicted (or initial assumption. The conceptual model defined) as the color that results from mixing resembles that of chemistry, which explains a red paint and yellow paint. The technical aspect large number of compounds as chemical of color theory covers questions about how combinations of a much smaller number of orange and other colors are mixed and behave in elements. What is a primary color? mixtures. The aesthetic aspect of color theory is Two definitions are given, neither an inquiry into the harmoniousness of orange satisfactory. The first says that the primary and other colors in combination with one colors are a set of colors from which all other another. We are told how to combine colors and colors can be mixed. The reasoning is a which colors to combine. variation on Aristotle's belief that colors are The technical side of color theory is a set mixtures of the opposites of darkness and light. of rules about how the gray scale, the color A problem arises with the first definition. wheel, and the color solid ought to be inter- Four-color process printing can create the preted, what each implies about color relation- illusion of iridescence or metallic hues. But we ships. The primary colors are an important cannot take red, yellow, and blue paints and stir concept common to modern color systems them together to create paints in silver and other

293 294 Primary, Secondary, and Tertiary Colors

metallic colors, dayglo colors, fluorescent In Principles of the Science of Colour, William colors, or iridescent colors. The argument that Benson, following Maxwell, identified the these colors are not really colors is not helpful. primary colors as red, green, and blue, linked Each visually excludes other colors, the with do, mi, so] in the musical scale (Benson criterion for identifying a color. 11868] 1930, 38). Benson proceeded to one of The second definition sidesteps the issue of those tortured comparisons between music (a how to mix silver and the other problem colors. product of human artifice) and color (a natural A primary color is defined as a color that cannot phenomenon) that were once thought to shed be mixed from others. This is dangerous ground. light on the nature of color. The practice of If silver cannot be mixed from other colors, as comparing colors and musical notes can be was the original problem, then silver must be traced to Plato, who regarded the homogeneous recognized as a primary color. This, I think, was quality of a musical tone extended in time, or of not the intent. a color extended in space, as examples of unity. By combining primary colors, secondary Both, therefore, were beautiful, as were colors and tertiary colors are created, accord- geometrical forms. ing to whether the mixture includes two pri- The English physician and amateur Egyp- mary colors or three. In theory, we proceed tologist Thomas Young (1773-1829) is often from a few basic units to every conceivable said to have provided a scientific foundation for color. the belief that a set of primary colors exists. The assumption that a set of primary colors Young, theorizing about color vision, reasoned exists can be traced to antiquity. Aristotle that each point on the retina cannot contain an reduced all colors to combinations of darkness infinite number of receptors, each attuned to one and light, effectively his set of primary colors, among an infinite number of colors. Each point though he did not use that label. Because must contain a limited number of receptors colors can be mixed from other colors, the keyed to a small set of basic primary colors. idea persisted that all colors could be mixed Colors other than these primaries are seen when from a small set of basic colors. What colors multiple receptors are stimulated in unison. ought to be included in the basic set has not Various ratios of stimulation account for the been clear. Leonardo da Vinci listed the primary entire range of perceptible colors. colors as black, white, blue, yellow, green, Young initially believed that red, yellow, umber, purple, and red (Rigaud 1957, 138). He and blue were the primary set. He later pre- contended, although few modern theorists ferred red, green, and purple. Fascination with would agree, that none of these colors can be primary colors lingers, promising to reduce a mixed from others. complex phenomenon to easily understand- The twentieth-century citizen asked to able terms. But Young's theory is obsolete identify the primary colors usually names red, today as an explanation of color vision. The yellow, and blue. Some people say red, yellow, set of color-keyed retinal receptors was never blue, and green. Others say red, green, and located. Modern research has identified two purple, or red, green, and blue. Those who basic receptor types: the rods and cones, exclude yellow do so on the reasoning that red named after their respective shapes. The rods, and green lights can be mixed to make yellow sensitive to differences in color value (light and light. Yellow therefore is not a primary color. dark) account for vision under conditions of Correlating the primary colors with other low illumination. The cones, sensitive to hue elementary units was a once popular pastime. or chroma, either cease to function or play a Primary, Secondary, and Tertiary Colors 295

subordinate role when little light is present. From the colors in any of these sets, all other The rods and cones, also called opsins, spectral colors can be mixed. contain several visual pigments, including This is not equivalent to demonstrating that rhodopsin (visual purple), porphyropsin, one set (rather than multiple sets) of lights car iodopsin, and cyanopsin. Each pigment be used to mix all colors (not just the spectra: bleaches in the presence of light. Absorption hues). Also, Young's belief that he had mixed maxima for rhodopsin and porphyropsin have all of the spectral colors flies in the face of been shown to differ from one animal species to Newton's findings. Newton contended that another. This may mean, as has been suggested, when rays were mixed to create green, this that all species of animals do not see the same green was not a color match for the singular colors. But the structure of the eye may not ray of spectral green. An unanswered question explain how and what people and animals see. is whether Young obtained different results or The rods of the human eye lie around the was just less discriminating in his observation periphery of the retina. The cones, necessary of colors. for color vision, are clustered near the center. Young's experiment suggests that the issue Yet we have no sense that color vision is better is spacing, not color. If all or most colors o1 the near the center of the eye. We sense no spectrum can be matched by mixing rays in any difference in acuity for color in different parts set of three properly spaced colors, the number of the visual field, although such differences of sets of primaries is too large for the word to may actually exist. have any meaning. None of these sets is Young is remembered primarily for two primary in the sense of being a unique set The experiments. In the first, light sent through two experiment was never properly assessed in the slits to a screen fell in an interference pattern of haste to use it as evidence that a set o1 primary alternating bands of light and shadow. This colors exists. We need to know why spacing is suggests light behaves like a wave, although important. Newton had regarded it as a stream of The logic of the usual argument from corpuscles. An adaptation of Young's Young's three lights leaves much to be desired double-split experiment became a cornerstone Assume that every spectral color can be mixed of quantum theory when it was shown that the from the primary colors in any set in which behavior of a beam of electrons, passing it is not included as a primary. Let A, B, and through one or both slits, could not be C be the colors in a primary set. Let X, Y, and explained by the laws of classical physics. Z be the colors in another primary set. If the Young based his theory of color vision on colors A, B, and C can be mixed in various the second experiment. He demonstrated that proportions to create every other spectral if three properly spaced light frequencies are color, each much be a primary. But A, B, and selected from the spectrum, all other spectral C are not primary colors if it is true that they colors can be created by mixtures of these can be mixed from X, Y, and Z, the colors in three. Various sets of colored lights can be another primary set. Young's experiment, said used. But each must contain no fewer than to show that a set of primary colors exists, is three lights. Young's experiment was per- better evidence that no set of primary colors formed with colored lights, which do not exists. Young showed that all or most spectral behave similarly to colored pigments. It pro- colors (not all colors) could be mixed from var- vides an experimental basis for the claim that ious sets of three colors (not a single set). To many three-color sets of lights are primary. press further is an improper generalization. 296 Primary, Secondary, and Tertiary Colors

Primaries Cannot Be Mixed from mixture that also results in red. Both Other Colors components are red, although neither is the A primary set of colors is not a necessary shade of red obtained as an end result. assumption, and the logic of the second Everyone knows that if two reds are mixed definition is no better than that of the first. together, the result is a third red, different from Assume that a primary color, say, red, cannot be either of its components although still red. This mixed from other colors. Let A and B be colors, is a genuine contradiction of the assertion that which may or may not be red. Let redl, red 21 no mixed color can be red. and red 3 be various shades of red, a color that is primary. Let 0 mean "cannot be mixed to make. " Secondary Colors The edifice of color theory, like a tower of B + A ≠ red1 Babel, rests on the wobbly assumption that a set A + red 2 ≠ red1 of primary colors can be isolated or is already red1 + red 2 ≠ red 3 known. The secondary colors, the next layer of elaboration, are obtained when two primary The first proposition is that red cannot be colors are mixed. If the primary colors are mixed from A and B, because a primary color assumed to be red, yellow, and blue, the cannot be mixed from other colors. Two secondaries are orange (red plus yellow), green propositions follow, according to whether A, B, (yellow plus blue), and purple (blue plus red). or both are red. Are any of the propositions Variation is available through adjusting true? proportions. A large amount of red and a small Although the first proposition is ensconced amount of yellow yields some shade of in color theory and popular belief, each redorange, an orange tending toward red. proposition is untrue. Colors can be mixed to Reversing the ratio yields yellow-oranges, form red. The only requirement is that one or oranges tending toward yellow. Proportion is more of the colors in the mixture be red. The critical in color mixture, though ignored in second proposition, a special case of the first, color theory. It could be taken into account in shows how to mix red. Imagine a large amount color notation by adapting the subscripts used of red paint to which a small amount of paint of in chemistry. In chemical notation, H20 is any other color is added. The red paint, though water and HZOZ is hydrogen peroxide. The not exactly the same as previously, will in most subscript numbers acknowledge that each cases remain red. To change red paint (or red molecule has a different ratio of hydrogen light) to another color, more is required than atoms to oxygen atoms. Following this method, admixture of a color other than red. The red to + yellow 3 might indicate a mixture of proportions of the mixture must be such that ten parts red, three parts yellow. Attention to redness is no longer visually dominant. If more proportion brings into focus the excessively and more of, say, black paint is added to red schematic nature of color theory. Indeed, red paint, the color changes to blackish red, a and blue make purple in some cases. But if a variety of red. The paint ceases being red at the tiny amount of red paint is added to a large point when so much black has been added that amount of blue paint, purple is not the result. we ought to classify it as reddish black (a shade For a second example, if the red is close to of black) rather than blackish red (a shade of orange, the blue is greenish, and the colors are red). mixed in equal amounts, most people will call The third proposition represents a type of the product color brown. Primary, Secondary, and Tertiary Colors 297

Tertiary Colors Color theory has nothing to say about mixtures The tertiary colors result when all three pri- of this sort, an unexplored frontier. mary colors are mixed. One way is to mix the Color variations based on surface sheen are primary colors, say, red, yellow, and blue, also excluded. Yet shiny black, for all practical directly. A second way is to mix any two purposes, is a different color from dull black. colors in the secondary (orange/green/purple) No object can be both shiny black and matte set. Because each secondary color is made black at the same time. Ultraviolet constitutes from two primary colors, a mixture of two another excluded class, though ultraviolet light secondary colors includes all three primary can be seen by some people. Because the light colors. can be seen, it can also be mixed with light of The tertiary range is usually identified as other colors. Paint that reflects ultraviolet brown, a simplification. It includes, depending wavelengths can be mixed with paint of other on the proportion of each primary component, colors. a large variety of russets, olives, browns, Black and white are excluded from the grayish browns, and bluish grays. The color primary/secondary/tertiary set. But black paint schemes in Rembrandt's paintings are based on and white paint are frequently mixed with paint a wide variety of tertiary colors. Some bluish in those colors. The colors the InterSaciety grays in the tertiary sets are so dark that they Color Council standardized as ISCC blackish approach black or near black. As a result, red and ISCC reddish black can be matched by people are sometimes uncertain about whether mixing black and red. Mixture of a primary, brown or black is the result of mixing paints of secondary, or with black yields all colors. colors described as grayed, darkened, or Surfaces absorb the colors (wavelengths) of blackish. More formally, these varieties of light that they do not reflect. The tertiary color are called shades or tones. colors, because they have more components, Mixed colors with a strong white component usually absorb more colors of light than are called pastels or tints. Distinctions are rarely primary or secondary colors. Brown surfaces, drawn, although they might be, between pastel which absorb many wavelengths (colors) of primary colors (say, pink), pastel secondaries light, and black, which absorbs all (pale green), and pastel ternaries (tan). The wavelengths, are the most common colors in omission suggests the coexistence, without the natural world. Nature's balance favors fraternization, of two separate sets of language conditions that allow the absorption or passage rules for naming colors. Some pastels have of light. The hues, as the least mixed, purest, names of their own: pink, tan, beige, eggshell, brightest, or most reflective of all colors, cream, , pistachio, mint, baby blue. suggest maximum deviation from this de facto Pastels that have a black component as well as a norm. We do not know, however, whether the white component are usually called grayed fascination of human beings with pure hues has pastels or whitened tones. a biological basis. All colors cannot be mixed from a set of pri- mary colors unless the primary/secondary/ter- Purity and Muddiness tiary set includes all colors. It does not. The ranking of colors as primary, secondary, Excluded ranges of color include metallic and tertiary is hierarchic, based on the pres- colors, fluorescent colors, iridescent colors, ence of one, two, or three primary compo- and dayglo colors. Paint in any of these colors nents. The hierarchy is limited to colors that can be mixed with paint of any other color. in theory can be reduced to primary colors. 298 Primary, Secondary, and Tertiary Colors

A color made by mixing, say, silver paint and of Anteus, whose mother was the earth, says he yellow paint has no place in the system. It could not survive unless he touched the dirt cannot be ranked. from time to time. When a person returning The linguistic tilt of the ranking system from abroad kneels to kiss the earth of the implies that the primary colors are primal or homeland, we understand what is meant and elemental. As the least mixed colors, they can regard it as very emotional. be regarded as pure. The tertiary colors are the Muddiness or dirtiness, when attributed to a most mixed or least pure. Following this rank- color, implies chromatic mixture: deviation ing, many idiomatic expressions refer in judg- from an ideal of chromatic purity that calls for mental terms to purity or lack of purity in fewer than three primary components. People individual colors. Browns, although many peo- who say they like purity in colors are rarely ple like them, may be called murky, muddy, discovered to mean pure brown or pure olive or dirty, rarely a term of admiration. When a green. Colors without spectral components such primary or secondary color is called murky or as silver and bronze are rarely called pure or muddy, this usually implies it tends toward impure colors. brown. It falls short of an expectation that Taste in color changes and sometimes is reds, yellows, blues, greens, and violets should volatile. Modern designers of clothing and be bright or "pure." A dark blue, say, navy home furnishings notice the difference between, blue, is rarely called muddy. The term is say, this year's olive green (which is reserved for blues with yellow and red fashionable) and last year's olive green (which components, blues that tend toward the tertiary is not). Long-term trends lie behind styles of range. the moment. Over the past forty years, people Why is muddiness bad in a color? I suspect have come to prefer bright colors, now widely the issue is human passion for order, expressed used in clothing, home decorating, office in a desire for cleanliness that goes back to an furniture, automobiles, pots and pans, and early date. The Old Testament sorts animals into bedsheets. The long-term change in taste was clean and unclean groups and forbids eating probably caused by the use of strong color in unclean animals, rules still followed today by modern art, or the influence on modern art of Orthodox Jews and Muslims. Unclean animals designers. Other twentieth-century influences are often associated with mud: the pig because it are American Indian and African textiles and wallows in mud, the carp because it gets its food the comic-strip colors of color television and from the mud at the bottom of ponds. Human video games. beings, we are told, are made of the dust of the Muted colors, like other colors, go in and out ground or of dirt, a humble origin. We were not of fashion. When fashionable, brownish blues made, as the angels may have been, from or olive greens are thought of as subtle rather heavenly or ethereal matter. than dull or dirty. The reinterpretation leaves As in the case of many popular ideas that the point of reference unchanged. A tertiary carry over into our judgments of colors, an color is called subtle in appreciation of its irrational bias runs through this dislike for subtle deviation from the norm of either the mud. Early human beings in Mesopotamia built primary colors or the spectral hues. their houses out of mud brick, a material used The reputation for purity of the primary since the New Stone Age. Clay, a variety of and secondary colors is as ancient as that of mud, is the basis of the ceramic arts. People the tertiary colors for lack of it. Again the roots might have praised this wonderful material, lie in religious metaphor. The primary colors, although the Bible does not. The Greek myth we say, cannot be made from mixture. Red, Primary, Secondary, and Tertiary Colors 299 yellow, and blue are pure in that sense. The unusually popular combination in Renaissance reputation of the hues for their own type of and Baroque painting, similarly relies on a set of purity comes partly from their brightness. colors in which each is distinctly different from Spectral green is brighter than olive green. the others. The cuts across Also, the spectra in which the hues appear are boundaries of style and time. Van Eyck used it created by light, pure because it is incorporeal. in The Man with the Red Turban (London, Divorced from the materiality associated with National Gallery), as did Rubens in Diana and the corporeal world, incorporeality-including Actaeon (New York, Metropolitan Museum of the incorporeality of light-suggests the realm of Art). the creator rather than that of his creations. Is this juxtaposing of strong hue with strong Although comets and eclipses of the sun value contrast inherently beautiful? Do we like might have seemed more amazing, thunder sets of colors in which, as in Mondrian's and the colors of the rainbow seem to have paintings or in the rainbow, each color looks been the events in the sky that most impressed very different from the others in the set? ancient peoples. Zeus, the Greek king of the Perhaps. But Rembrandt, considered a great gods, carried a thunderbolt. In India, the colorist, offered tertiary schemes built on small rainbow and the thunderbolt were symbols of color variations within a limited range. The Indra. The Bible said the rainbow was set in the appeal of Rembrandt's colors lies in the sky after the flood as a sign of God's covenant similarity of one color to another. Rembrandt with human beings. The rainbow is a familiar and Mondrian are often called spiritual, an figure for purity or resurrection, and the search effect attributed to the colors in their paintings. for purity, or mourning for its loss, is a potent Yet Mondrian used unmixed primary and theme in Western literature. achromatic colors. Rembrandt used a wide In the visual arts, rainbows and other light range of subtle tertiary variations. effects are popular in painting from the late Why do certain colors or combinations of Renaissance through the nineteenth century. colors create a sense of religious awe? Why did But the idea that some colors are more pure Mondrian's interest in the spiritual lead him to than others was not forgotten in the twentieth simple colors? I suspect the issue is speechless- century. The use of color in the paintings of ness, the condition of having no words to Piet Mondrian turns on an association of purity describe what we see. We have no idea how to with "pure colors." Mondrian's search for an art translate into words the nature of blueness or its that was to be spiritual or pure and that, for differentness from redness or whiteness. consistency, had to be created with pure colors, Speechlessness by its nature approaches led to a gradual paring of his palette. The religious awe, as if the inexplicability of color pinks, browns, and other mixed colors found in were an objective correlative for the Mondrian's early paintings gave way to a inexplicability of the absolute. palette limited to primary and achromatic Much earlier than the paintings of Mon- colors: red, yellow, blue, black, white. No drian, we can turn to Sumerian and Byzantine other colors, apparently, were sufficiently art to see color used for religious purposes, spiritual, pure, or elemental. though not the colors Mondrian chose. The gorgeousness of Mondrian's combina- Sumerian and Byzantine art, both from the same tion of red, yellow, blue, black, and white part of the world, relied heavily on the seems to come from the uniqueness of each gorgeousness of gold leaf and silver leaf, color, different from any other in that set. Red, materials with distinctive colors. After the Mid- white, and black (or red, light, and dark), an dle Ages, these and other metallic colors were 300 Primary, Secondary, and Tertiary Colors

rarely seen in painting until the 1960s. Gold manner. and silver came to be thought of as not really Set 1: A, B, C (red, yellow, blue) colors and went unmentioned in books that Set 2: AB, AC, BC (orange, purple, green) advised on how to combine and mix colors. Set 3: ABC (brown)

Our choice is whether to assign the greater Turning the System Upside Down value to the extreme of the scale which is most Hierarchical rankings that are disguised value complex or that which is least so. judgments can be turned upside down without Why do we feel compelled to reason in gain or loss of interior logic. We need not either of these judgmental manners? The boring assume that complex substances, say, the ter- exercise of defining one color in terms of tiary colors, are less pure than substances that another, like the boring exercise of defining are simplex, say, the primary colors. Imagine women in terms of their differences from men, reversing the system. The complex units, the interferes with what might otherwise be a less tertiary colors, shall be the norm. self-deceiving vision of the nature of the Rather than grading colors from pure to experiential world. Red, blue, brown, silver, impure, we shall rank them from complete to dayglo green, yellow, pale lavender, and all incomplete. Brown, because it includes all the other colors are each equally unique, each best primaries, will be described as a complete appreciated for its individual qualities. color. Red, because it does not, will be called In the type of metaphor passed along to us incomplete. A reversal in public taste would by earlier peoples, the world, divided, is a occur. People would assume that the brighter series of fragments to be reassembled into hier- variations of color were distressingly incom- archies, each resembling a king ruling over the plete rather than gratifyingly pure. The con- subjects beneath him. We need more refined ventions of language would encourage the new metaphors, less autocratic images. A better assumptions. exercise is to imagine dividing a blue spot of A grading from complete to incomplete is color in half. This division does not make the no better or worse than a ranking from pure to spot less blue and suggests that the whole is impure. Either makes sense if ranking is to be inherent in every part. All colors are beautiful, according to the number of components. in the sense that all people are beautiful. This is Assume that red, yellow, and blue are coded as not the message of color theory, with its A, B, and C. The primary, secondary, and rankings that are rarely either logical or con- tertiary colors can be notated in the following sistent with perceptual experience. CHAPTER 33

Complementary Colors

Each color is closely related to another called its complement. Walter Sargent, The Enjoyment and Use of Color In pigments, complementary colors mixed together will cancel each other out, creating a neutral tone of gray or black. Joshua C. Taylor, Learning to Look

n logic the complementary class to red can mixture of the other two primaries. Color be identified as not-red. In color theory wheels are arranged so that each primary lies com plementary has a specialized meaning opposite its complement. The concept can be and applies to the sets below. One color in extended to any hue variation with appropriate each set is a primary color. Its pair is the shifts. The complement of an extremely secondary color made by mixing the two yellowish red is an extremely bluish green. As remaining primary colors. the balance of the red shifts toward yellow, that of its complement rotates in an opposite • red and green (green is yellow direction. and blue) This traditional arrangement of color • blue and orange (orange is yellow wheels implies that an equal number of inter- and red) mediaries lies between any two major hues. Yet • yellow and (purple is red we do not know whether the number of, say, purple and blue) blue-greens the average observer can dis- tinguish is matched by an equal number of In a three-primary system, here assumed to yellow-oranges. Symmetry in distribution consist of red, yellow, and blue, the comple- seems unlikely. The eye shows variations in ment to any primary is the secondary made by sensitivity and discriminates with greater fine-

301 302 Complementary Colors

ness in the yellow-green range. The symmetry component. No answer can be given to the implied by the color wheel probably does not question of what colors are the complements to exist and is never the same symmetry with any silver and bronze. The tertiaries are bypassed, two theorists. with less logic. If any given brown is to be The Ostwald wheel has eight basic colors. regarded as some ratio of red, yellow, and blue, Purple and leaf green (more yellowish than sea we can ask that an opposite ratio be identified. green) occupy 25 percent of the hue circle. The We can also ask what color (presumably arrangement implies, though Ostwald may not another brown) corresponds to this opposite have thought about it, that 25 percent of ratio. distinguishable hue variations lies in this sector. The question of whether complementarity On the ten-sector Munsell wheel, 40 percent is can exist among grays, even though not occupied by the colors that look most similar: between black and white, turns on the distinc- red-purple and purple (together corresponding tion between achromatic grays (made of black to Ostwald's purple), and greenyellow and and white only) and those tinged with hue. yellow (corresponding to leaf green). What Among the achromatics, light gray is not called percentage of visually distinguishable hue the complement of dark gray. But theorists variations lies in the yellow-green through usually agree that the complement of a reddish purple range? Is it 25 percent (Ostwald) or 40 gray is a greenish gray. The Ostwald color solid percent (Munsell)? Without knowing, we cannot is constructed to compel this identification. determine whether each color has been paired The universe of discourse for complemen- with its correct complement. tarity is the hue component of the set of pri- Treated at some length in Helmholtz's mary colors. Yellow is called the complement Physiological Optics, the concept of color of purple because the issue in complement- complementarity-that each color can be paired arity is hue rather than value (lightness or with an opposite color-is widely assumed to darkness). We are asked, in making the have a rational or scientific foundation. Yet it identification, to ignore the lightness of yellow can have no such foundation unless a unique set and the darkness of purple. Even though of primary colors can be identified. This lightness and darkness are said to be opposites, primary set must include one exact shade or this variety of oppositeness is not considered in wavelength of, say, red, not just the range of this case. reds generally. Injunctions to learn the complementary Carried over into the concept of color com- colors abound, which presupposes a value in plementarity is the familiar baggage of miscon- doing so. For Sargent, "one who uses colors ceptions about color that have persisted for should know the exact hues of the more centuries. Complementarity does not apply to important pairs of pigment complements" those colors, say, black, that are assumed to (Sargent [1923] 196, 115). For Bragdon, be not really colors. Complementarity is to be "complementary colors should be learned and understood as a function of hue, of the primary committed to memory, just as a musician and secondary colors, or of sets of those recognizes and remembers consonant musical colors. Black and white are not convention- tones" (Bragdon 1932, 117). Guptill "cannot ally included in a listing of complementary over-emphasize the importance of these sets, though black is often called the opposite complementary hues in work with pigments. of white, and "no light rays" is defensibly the You should memorize the principal pairs" logical complement to "all light rays." Simi- (Guptill n.d., 87). larly excluded are all colors that lack a hue Like the admonition to study opposites, the Complementary Colors 303 dicta are impossible to follow. The particular Figure 33-1. The complementary colors identified by five theorists. Given the absence of firm criteria for shades of color opposite one another are never complementarity, the colors identified as the same from one color theorist's wheel to the complements vary from one theorist's wheel to another's. And similar names need not imply similar next. This leads to too many answers to the colors. (Complementaries from the color wheel question of what the complements are. Too few devised by Fritz Trautman after Bragdon 1932, 123. Ross complementaries after Ross 1919, 3.) answers exist to that other question of how the determinations were made. Ranking colors in OSTWALD (8 major hues) red and seagreen pairs of opposites sows confusion and leads to orange and turquoise blue no useful insights. yellow and ultramarine leaf green and purple , asked to define Impressionism, identified two principles: "study MUNSELL (10 major hues) red and blue-green the complementaries" and "don't mix yellow-red and blue pigments." Pissarro's color system, diagrammed yellow and purple-blue green-yellow and purple by Brown (1950), included ten major classes of green and red-purple hue grouped in five complementary pairs: BIRREN (12 major hues) • red and blue-green red and green red-orange and blue-green • yellow and blue-violet orange and blue • green and red-violet yellow-orange and blue-violet yellow and violet • blue and orange yellow-green and red-violet • violet and yellow-green ROSS (12 major colors) red and green The number of alternative arrangements is as red-orange and green-blue large as the number of theorists offering orange and blue orange-yellow and blue violet opinions. Figure 33-1 shows the complemen- yellow and violet tary pairs according to Munsell, Ostwald, yellow-green and violet-red

Birren, Ross (who followed Chevreul), and TRAUTMAN (16 major hues) Trautman. If asked to name the complement redness and blueness red-orange and blue-indigo of blue-green, the respondent should reply orange and indigo "red" according to Pissarro or Munsell. On orange-yellow and indigo-violet yellow and violet Birren's wheel, the answer is red-orange. For yellow-green and violet-purple Trautman, the color is green-blue rather than green and purple blue-green; its complement is purple-red. The green-blue and purple-red complement of yellow, according to the theorist, is variously blue-violet, purple-blue, needed on more than the principle that colors ultramarine, or violet. The differences between have complements. The perennial question systems reach beyond naming to the placement about complementarity is what the term means of hues. in relation to color. Unless it can be linked to For Birren all theorists were equally right, perception-correlated with something people not equally wrong, because "generalities alone see-complementarity remains a conceit, a label are important," and "whether yellow finds its that provokes endless argument because it has true complement in ultramarine blue (Ost- no intelligible meaning. wald), in purple-blue (Munsell) or as [ in Bir- ren's system] is of no great consequence" Complementarity as Negation (Birren 1969, 81). But if color complementar- What can I see in two colors that shows me ity is a meaningful concept, agreement is they look opposite to one another? A common 304 Complementary Colors

answer is that complementary colors negate disc. Whether each paper should cover half the one another. This negating is not the same as disc's surface is unclear. In one variation on mutual exclusivity. Although the blue I see the theory, when the disc is spun, it ought to might be said to have negated the red I cannot look gray if A and B are true complements. The see at the same time and place, this leads to no mutual negating is a negating of chromaticity, more than the conclusion that any color is a which in this case results in an achromatic complement to all colors other than itself. gray. Evidently mixtures of colors can be ranked Given that every theorist structures the color according to whether negation occurs. Blue and wheel with that theorist's chosen set of orange, because complements, would negate complements, two outcomes are possible: most one another when mixed. Blue and yellow sets will not pass the test, or the test is (noncomplements) would not. Most unreliable. If the idea of color complementar- developments of the negation theory specify a ity has a meaning (it may not), five or more third color to be obtained when complements different shades of orange cannot each be are mixed. What color? identified as the exact complement of a A mixture of two complementary colors particular shade of turquoise blue. If a + b = c, includes all three primary colors. Thus, we and a and c are known, multiple values do not would expect the mixture to look brown, exist for b. whether or not the same shade of brown in each Claims of ability to pass the disc test are case. In the several explanations of comple- common, and at odds on critical details. We are mentarity, however, the product color is not told what proportion of each color should be variously identified as brown, white, gray, or used, or whether the same value of gray results black. We need more information about this in each case. If red and green spin to achromatic product color, the criterion for whether a gray, and yellow and purple also do so, will the mixture of two colors causes them to negate one gray be equally light or dark in each case? another, whether the colors are complements. Bitten contended that if complementary Munsell colors are "combined in equal 50-50 percent proportions and spun on a motor, the resultant The Maxwell Disc Test visual mixture would be a neutral gray" One of the more technologically ornate (Munsell 1969, 54). T. M. Cleland, an associate explanations of negation holds that the colors in of Munsell, required that the colors be mixed in any complementary pair spin to gray on a proportion to their chroma (Munsell 1969, 28). Maxwell disc or color top. James Clerk The prescription is so vague as to be Maxwell invented the device, popular for many meaningless, and it is not the recipe Bitten years in the schools. The Maxwell disc is a disc advised for spinning Munsell colors. Ostwald on a spindle caused to revolve by a motor or claimed the complements identified on his other mechanical means. Maxwell's original wheel spin to "yield a neutral gray," though in instructions for making one have been updated unspecified proportions (Ostwald [ 1916] 1969, by later authors (Maxwell [ 1890 ] 1965; 35). Jacobson 1948; Smith 1965). A modern Smith used Maxwell discs to assess the Maxwell disc can be made of cardboard, cut to accuracy of the eight Shiva maximum com- the size and shape of a phonograph record, and plementary colors. These paints were "de- spun on the turntable of a record player. signed to mix in equal proportion by weight To test colors A and B for complementar- to a middle gray in pairs of complements," ity, papers of these colors are fastened to the though the relative weight of a pigment or Complementary Colors 305 paint has no known correlation with its color neutral gray. When, however, we spin the top (Smith 1965, 91). No pair passed the test. One with violet and yellow disks no adjustment of of the four pairs spun to a violet-gray, another their proportions will give us an exact neutral. to a yellowish gray. The other two failed to The gray will be slightly pinkish .... Blue and spin to the same gray. The ratio of color areas yellow disks produce gray when rotated. In required to produce any gray was never 50 pigments, however, blue and yellow produce percent/50 percent. It ranged from 21/79 green . . . . In order to produce gray with paints, percent to 30/70 percent. Smith concluded that the color which must be mixed with blue is although Shiva's maximum complementary orange, but with the color top the blue and colors "show a highly unsatisfactory orange disks give a violet purple" (Sargent performance record when mixed on the [19231 1966, 111). Maxwell wheel," they are a "set of true Maxwell believed that the color "produced complements" (Smith 1965, 91). Is the problem by fast spinning [of a Maxwell disc] is identical the Maxwell disc mixture, the maximum with that produced by causing the light of the complementary colors, or both? different colours to appear on the retina at once" Guptill was not enthusiastic about Maxwell (figure 33-4). This is unlikely in light of what discs, saying "this method of color mixture we know today. Color results obtained by disc was for many years more confusing than help- mixing are not those predicted or obtained from ful. Although the fascinating little device either light or pigments. The factors that blended colors to produce innumerable hues, probably account for the disparity are multiple it never quite proved out with either the reflectance and metamerism, a visual similarity Brewster or Young-Helmholtz theories of between colors with dissimilar sources or color mixture" (Guptill n.d., 18). Listings from constituents. Pigments and dyes, including those Cholet indicate that colors mix on a Maxwell used to manufacture colored papers, usually disc in a manner inconsistent with any theory reflect more than one wavelength of light. Even about mixtures of either colored pigments or if the color of a piece of red paper visually colored lights (figures 33-2 and 33-3). Sargent resembles that of a beam of monochromatic reported that, "in fact, we find that the color light of a specific wavelength, the paper is which with another produces gray when unlikely to be reflecting light of only that mixed with paints, is seldom if ever the same wavelength. The beam and the paper cannot be hue that produces gray with it when the mix- regarded as equivalents. ture is made with disks, on the color-top, or The Benham disc, a cousin of the Maxwell with colored light .... When we mix violet disc, is black and white while at rest but red and yellow pigments we can approximate a and blue in motion. The behavior of the Ben-

Figure 33-2. Comparative results of Maxwell disc mixture and pigment mixture. Maxwell believed that mixing colors on spinning discs could be equated with mixing beams of colored light. But this is not the case. Results in Maxwell disc mixture are also typically dissimilar to results from mixing pigments. (After Cholet 1953, 22).

306 Complementary Colors

Figure 33-3. Mixtures on Maxwell discs required to match mixtures on palette. Colors obtained by mixing on d can be correlated only uneasily with those from mixing pigments and require different proportions. (After Chc 1953, 22.)

ham and Maxwell discs suggests that hue can be of any mixture including three primaries, created, modified, or destroyed by motion, a irrespective of whether these primaries were phenomenon unexplained by any present theory ever arranged to form a pair of complementary of color. Would the color seen on a disc colors. Because a tertiary color can be made spinning at a particular velocity be modified if from two secondary colors, brown is the the disc were accelerated or slowed down? We predicted result of mixing, say, orange and do not know. Like the slit phenomena explored green, orange and purple, or purple and green. by Thomas Young, disc phenomena are little None of these pairs are complements. understood. If black, gray, or any color other than brown, this conflicts with the definition of a tertiary color. Mixtures of complements The Product Color Test include all three primaries, and are cones- Whether Maxwell discs perform as predicted is quently predicted to produce brown. They of small interest except for the issue of com- cannot at the same time produce gray or black. plementarity. If complementary colors cannot The gist of the inconsistencies is that brown be shown to negate one another by spinning cannot be a test of complementarity, yet must to gray on the discs, how can we identify pairs be the test. The color is the predicted end of complementary colors? Variant versions of result when any mixture includes all three the theory do not rely on Maxwell discs. They primaries. suggest colors other than gray to be obtained This leads to a negative test of sorts: if a under various circumstances. Complementar- mixture of two colors does not produce brown, ity between two colored lights is said to be the colors are not complements. We must proved if they can be mixed to form white therefore eliminate black or gray. Mixing two light. Proportions for the mixture are not pigments to produce black or gray proves the specified, nor is it clear whether the white will colors are not complements. Brown is the end be identical with that of a beam of natural white result of mixing complements, although also an light. end result of mixing various color pairs that are Mixtures of pigments of complementary not complements. colors are said to form, variously, brown, gray, To test the logic of the various theories, let or black. Proportions, again, are unspecified. A, B, and C be primary colors. Let ABC be If brown, this is not a test of complementar- brown, the predicted color of any mixture that ity. Brown is elsewhere identified as the color includes A, B, and C. Let AB be the secondary Complementary Colors 307

color complementary to primary color C. Let Other Tests Z be any color other than brown. The defini- Other explanations of how to test for com- tion of a tertiary color then implies the follow- plementarity verge on the incoherent. Fuchs ing about any pair of complementary colors. offered the thought that "the complementary colors are mutually destructive, each exclud- AB + C = ABC ing the other from sensation" (Fuchs [ 1908 ] Z ≠ ABC 1924, 242). Ostwald put his faith in the "law AB + C ≠ Z of least resemblance." This is said to determine The result of mixing complementary colors that "there exists for every hue in the hue cir- must be brown and cannot be any other color. cle another that is most different from it, which Yet brown is not a test for complementarity, is called its complement" (Ostwald [ 1916] because the color is predicted for any mixture 1969, 34). That less resemblance exists including three primaries. Mixtures including between, say, yellow and purple than between three primaries exist that are not mixtures of yellow and red is largely a matter of opinion. complementary colors. Ostwald provided no further information about the workings of this convenient, AB + BC = ABC elsewhere undocumented, law. BC + AC = ABC Sargent reported, "by looking at the AC + AB = ABC afterimages of all our (colored) disks we can A +B + C + ABC

Figure 33-4. Construction of Maxwell discs. Maxwell discs, in their day, were popular in both the schools and scientific laboratories.(From Maxwell [ 1890] 1965, 122.)

The colored paper is cut into the form of discs, each with The two combinations being close together, may be a small hole in the center, and divided along a radius, so as accurately compared, and when they are made sensibly to admit of several of them being placed on the same axis, identical, the proportions of the different colours in each is so that part of each is exposed. By slipping one disc over registered, and the results equated. another, we can expose any given portion of each colour. These equations, in the case of ordinary vision, are These discs are placed on a little top or teetotum,consisting always between four colours, not including black. of a flat disc of tin-plate and a vertical axis of ivory. This From them, by a very simple rule, the different colours axis passes through the center of the discs, and the quantity and compounds have their places assigned on the triangle of of each colour exposed is measured by a graduation on the colours. The rule for finding the position is this:-Assume rim of the disc, which is divided into 100 parts. any three points as the positions of your three standard By spinning the top, each colour is presented to the eye colours, whatever they are; then form an equation between for a time proportional to the angle of the sector exposed, the three standard colours, the given colour and black, by and I have found, by independent experiments, that the arranging these colours on the inner and outer circles so as colour produced by fast spinning is identical with that to produce an identity when spun. Bring the given colour to produced by causing the light of the different colours to fall the left-hand side of the equation, and the three standard on the retina at once. colours to the right hand, leaving out black, then the By properly arranging the discs, any given colour may position of the given colour is the centre of gravity of the be imitated and afterwards registered by the graduation on three masses, whose weights are as the number of degrees the rim of the top. The principal use of the top is to obtain of each of the standard colours, taken positive or negative as colour-equations. These are got by producing, by two the case may be. different combinations of colours, the same mixed tint. For In this way, the triangle of colours may be constructed this purpose there is another set of discs, half the diameter by scale and compass from experiments on ordinary vision. of the others, which lie above them, and by which the second combination of colours is formed. 308 Complementary Colors

find the color that is the complement of each" determining which colors correspond to these (Sargent [ 1923 ] 1966, 104). The technique is wavelengths. If the colors cannot be firmly not reliable. Afterimages are too faint to allow identified or are inconsistently named and precise discriminations of hue to be made. A identified among observers, the assertion that Color Dictionary (The North American Society the wavelengths are those of complementary of Arts, Inc., 1931, unpaginated) reported that colors has no foundation. "as two opposite colors (such as blue and Delving more deeply into the technical orange) are brought together, one neutralizes the literature does not resolve the inconsistencies, a other until the brightness or chroma of each pervasive condition in color theory. Helmholtz, changes from the most intense to the dullest followed by Arnheim, identified wavelengths of stage." 607.7 and 489.7 millimicrons (respectively, Beyond the circus of capriciousness yellow-orange and blue) as an exhibited by color systems and their achromaticity-creating pair of complements champions, experiential reasons can be cited (figure 33-5). Kries, Frey, Klinig, Dieterici, for wondering whether color complementarity Angier, and Trendelenburg disagreed, as does is a meaningful concept. No child can be taught nearly every other authority (figure 33-6). The ostensively that red looks more different from Optical Society of America found that the green than from black, yellow, blue, white, complementary wavelength to 489.7 is 605, not silver, or any other color. Ostwald's law of least 607.7 (figure 33-7). Ostwald's diagram of resemblance refuses to make itself manifest in complementary wavelengths indicates that 482 any visual form that can be exhibited. It can or 483 (not 489.7) is complementary to 607.7 only be communicated as a decree or accepted (figure 33-8). as dogma. The difference between Helmholtz and Ostwald looms larger than the small numeri- cal discrepancy suggests (see figure 33-8). Complementary Wavelengths in Light Between three and four of the twenty-four Ost- Arnheim noted the "Babylonian confusion that wald hue steps are involved, 15 percent of that reigns once we pass beyond what is said to be circle. Further, according to Ostwald's com- verifiable: that mixtures of certain spectral putation, 489.7 millimicrons is one of several colors are said to create an impression of wavelengths devoid of any complement, a achromaticity for the average observer," condition indicated by the gap on the right effectively, that complementary wavelengths of side of Ostwald's hue circle. A wavelength of light mix to form white (Arnheim 1956, 348). 607.7 millimicrons corresponds to yellow- Is even this verifiable? One difficulty is in orange or orange, depending on the theorist.

Figure 33-5. Complementary wavelengths in light. Because mixtures of light give results that differ from mixtures of pigments, complementary wavelengths in light cannot be correlated with complementary colors in pigments. (After Helmholtz [1909] 1962, 2:126.)

Color

310 Complementary Colors

from 433 to 656.2 millimicrons. The com- plementary wavelengths Helmholtz identified show ratios from 1.19 to 1.334, within a possible range of approximately 1.00 to 1.52 (656.2/433 = 1.52). In the technical literature, as in popular works and school texts, the persistent issue is verifiability. How Ostwald determined that the complement to a wavelength of 607.7 mil- limicrons was 482 or 483 millimicrons is unclear. Nor do we know how others arrived at other numbers. The complementary sets in the Ostwald system are dissimilar to those shown by Munsell or by anyone else. Each arrangement impressed its advocates as correct, though we are not told why. Figure 33-8. The Ostwald complements. The Ostwald complements differ from those of Helmholtz and the observers he cited. Colors in the red-violet range are shown without assigned Munsell and Ostwald Complementaries wavelengths. Inches can be translated into centimeters, A wavelength of 489.7 probably falls in the although the two cannot be mixed in the same blue range but is identified in some systems as computation. The skewing between the color blue-violet. wheel of one theorist and that of another is of A test for complementarity that is indepen- such a nature that no comparable form of dent of the color obtained by mixture is noted translation is possible. The Munsell hue circle, primarily in the technical literature. As Helm- divided into ten basic colors that have names, holtz and Arnheim observed, complementary uses a decimal system to provide a total of 100 wavelengths are said to have a roughly simi- hues. The order of the spectral colors (red to lar ratio: approximately 1.25 (Arnheim 1956, violet) reads clockwise. In the traditional 348). This cannot mean, however, that the manner, colors thought to be complements lie complement of a wavelength can be deter- opposite one another. mined by computation. Given lack of agree- The Ostwald hue circle includes eight basic ment on the particulars of which wavelength (or colors with names. Two intermediate colors which color) complements which, how a placed between each provide a total of twenty- reliable table of ratios can be developed is four hues. These are numbered from one to unclear. twenty-four, beginning with yellow. The order Whether the 1.25 ratio is even significant of the spectral colors (red to violet) reads coun- is debatable. According to Ostwald, the terclockwise (For the Munsell wheel, see Mun- wavelengths for visible light range from about sell (1905] 1961; 1969, 82. For the Ostwald 420 to 660 millimicrons, a narrow range (see wheel, Jacobson 1948, 26; Ostwald [1916] figure 33-8). Between these extremes, no two 1969, 83). Ostwald identified three hundred numbers can have a ratio to one another of less hues as an absolute maximum that can be than 1.00 (660/660 = 1) or more than 1.58 included in a hue circle. Beyond that number, (660/420 = 1.571). According to Helmholtz, differences in steps are held to be so minute the range of wavelengths for visible light runs as to be no longer visually discernible. In the Complementary Colors 311

Ostwald circle, as in Munsell's, "perfect vis- color. Although yellow and blue (hues 1 and ual complements lie opposite each other .... 13) are complements in the Ostwald system, Such pairs will cancel into gray if mixed on Munsell yellow and blue are not. Nor is a [Maxwell disc]" ( Ostwald [1916] 1969, Munsell's blue the same color as Ostwald's. 84). Munsell blue is closest to the Ostwald color In books, different printings of Ostwald and called turquoise. The color Ostwald calls blue is Munsell hue circles show noticeable color identified by Munsell as purple-blue. variations and an inferior level of color control Colors opposite one another on one wheel by industrial standards. Noticeable color are in many cases not opposite on the other. differences are more likely to occur in two Munsell's blue and yellow-red are complements. printings of the Munsell hue circle than in two The question of which Ostwald colors look most American flags made by different similar to them is complicated by color manufacturers (Compare Munsell hue circles in variations in different reproductions of the Munsell 1969 and the Munsell Company's Ostwald wheel. The closest match is probably Munsell Student Chart; Ostwald hue circles in either Ostwald's turquoise 16 and orange 4 ( Jacobson, 1948 and Ostwald [ 1916] 1969). Jacobson 1948) or turquoise 16 and yellow 3 Adjusting for this limitation of commercial (Ostwald [1916] 1969). The Ostwald pairs printing, the sets Munsell identified as appear to be one step off of complementarily. complements cannot be reconciled with the Munsell's red and blue-green are Ostwald complements in either color or name. complements. The closest match in Ostwald's As previously noted, on Munsell's wheel, red, wheel are red 7 and turquoise 17. The Ostwald orange (called yellow-red), and yellow occupy colors are indicated to be two steps away from 30 percent of the circle. On Ostwald's, these complementarily (8 percent). Munsell's colors occupy 37.5 percent, a difference of 25 red-purple and green, a complementary pair, percent. Another 25 percent of the Ostwald are difficult to correlate with any of the colors wheel is occupied by the color classes leaf on Ostwald's wheel. The closest match may be green and purple. Munsell's corresponding Ostwald's red 9 and leaf green 22, one step off colors, yellow and green-yellow (corresponding of the complement. Munsell purple and to leaf green) and purple and red-purple greenyellow are complements. The closest (corresponding to purple) occupy 40 percent, a match appears to be Ostwald purple 11 and leaf difference of 60 percent. green 24, one step off of the complement. Among the consequences of this skewing, both theorists may show the same pair of colors as complements, designated by dissimilar Complementarily and Harmony names. Munsell identified as complements Whether or not a meaningless concept, color colors he labeled yellow and purple-blue. Two complementarily has a rich mythos. For Bir- colors that look similar to this pair appear in ren, "a number of studies in the field of psy- Ostwald's circle, similarly as complements. The chology have verified the observations of Ostwald colors, however, are labeled yellow Chevreul that colors look best (a) when they and blue, respectively hues 1 and 13. are closely related or analogous or (b) when Some colors look so dissimilar that no com- they are complementary or in strong contrast" parison can be made. Examples include blue (Birren 1969, 35). Elsewhere Chevreul, who and the green with the highest yellow compo- questioned whether complements could be nent. The same name may be used for dissimi- identified, was credited with discovering that lar colors, or different names for the same "the complementary assortment ( of colors) is 312 Complementary Colors

superior to every other" (Birren 1969, 38). demonstrating that, say, a particular red is the For Jacobson, complementarity headed a list primary red, meeting conditions that cannot be of six modes of color harmony ( Jacobson 1948, met by any randomly selected bright red. Until a 56). Analyzing twelve paintings, he found that primary set can be identified or shown "the amazing accuracy of the complementary necessarily to exist, no foundation exists for the colors in these paintings is a final indication that conception of absolute secondaries or great artists are almost as sensitive to color complements. differences as modern spectrophotometric The elusiveness of primaries, secondaries, instruments" ( Jacobson 1948, 56). Color and complements accounts for the differences discriminations made by eyes, which need not between Munsell's system and Ostwald's and be the eyes of great artists, are more refined for the lack of agreement between any two than those made by instruments (Evans 1948, theorists on any point other than the impor- 203). Beyond this, Jacobson's criterion for tance of color theory. It accounts for the off- accuracy is unclear. Colors opposite one another set between one theorist's wheel and the next, on Ostwald's wheel, because believed to be a skewing that ensures the exact green oppo- complements, will not lie opposite one another site a particular red will not be the same in any on Munsell's. two systems. Bitten's argument that all color As Chevreul pointed out, color com- systems are equally correct is not convincing. plementarity falls by the wayside unless abso- I conclude the systems are all equally in- lute primaries can be identified. This requires correct. CHAPTER 34 Color Mixture

Mixtures of blue and yellow as lights produce white. Blue and yellow [Maxwell discs produce gray when rotated. In pigments, however, blue and yellow produce green.

Walter Sargent, The Enjoyment and Use of Color

fixed colors look similar to their task by eye, with no exact knowledge of how components in some cases, though the original paint was mixed. not always. Reddish gray resembles The answers to questions about color were red and gray. Green resembles neither blue nor once sought by studying music (figure 34-1). yellow. We cannot always judge, by looking at In a reversal used to explain color mixture, we colors, how they will look when mixed. The are often asked to consider differences color wheel allows results to be predicted between musical chords and mixed colors, as if approximately, but not with precision, the those differences were not to be assumed. limited truth behind the truism that blue and Maxwell mused that in a mixed color "we can- yellow make green. Nobody can select, from a not directly recognize the elementary sensa- chart of thousands of shades of green, the exact tions of which it is composed, as we can shade that will be produced by mixing four distinguish the component notes of a musical parts Prussian blue, one part cadmium yellow, chord" (Maxwell [1890] 195, 2:271). Gregory and two parts white. Human beings work concurred (Gregory 1966> 119). For Wright, effectively when assessing a mixed color they "there is nothing corresponding to [color mix- see, less effectively guessing the colors of mix- ture] for sound, since the ear does not synthe- tures not available for viewing. Given paint to size the aural response in the way the eye match, the trained person accomplishes the integrates the visual response. When a chord

313 314 Color Mixture

Figure 34-1. Ophthalmic color scale correlating the criterion, the person can be asked to mix colors with musical notes, developed by Louis colors to match, say, a particular green. Wilson. Wilson's scale is apparently based on the assumption that colors will be pleasant to look at if Unfortunately, if the task is accomplished, we the musical notes to which they are keyed create a are not free to conclude that the components of musical combination that is pleasant to hear. (After the original mixture were correctly identified. A color can be matched without knowing its components. Many ways exist to replicate any mixed color, the phenomenon colorimetrists call metamerism. A medium gray can be made by mixing black with white. It can also be made by mixing light gray with dark gray. A set of yellow-greens that match one another can be obtained by mixing various proportions of yellow with blue, blue-green, or green. What is indeterminable from looking at a mixed color is which of a range of possible combinations of music is struck, the listener, especially if he accounts for the mixture. Using any is trained, can identify the individual notes combination in the range, the color can be comprising the chord" (Wright [1944] 1969, replicated. The nearest parallel is number. Ten 69). is the sum of nine and one, and also the sum of The reasoning, by now, is pro forma. We six and four. should not need to be told that strong Except in laboratory experiments, laser similarities rarely occur between systems shows, and theatrical lighting, mixing colored created by human beings (say, music) and lights is an infrequent activity. Mixing colors natural phenomena (say, color). Musicians to match other colors usually involves mixing identify the notes in a chord because they know paints or dyes. The task, mysterious to the music, which has a limited range and strict rules uninitiated, is a commonplace skill on which about what sounds can be used and how they industry routinely relies. Mixing and matching can be combined. No amount of training colors, like drawing the human figure, can be enables anyone to separate the components in a taught to almost anyone,usually in a brief time. mixture of sounds, as in the merging of voices Although not as confusing as comparisons heard as the roar of a crowd. between color and music, those between color Nontraditional music is similarly resistant to and sound are similarly limited. Optical mix- instant analysis, for it more nearly approaches tures differ from aural mixtures. Sounds, the complexity of combinations of natural including musical notes, are considered mixed sounds. Rarely can one listen to if the components occur at the same time in computer-generated music and deduce the near proximity to one another. The location program by which it was generated or the from which individual sounds originate is sounds that went into the mixture if these irrelevant, provided the sounds can be heard. sounds were not conventional musical notes. The same chord has been played whether one Although the components cannot be musician strikes its three notes or three musi- identified in a mixture of colored lights, cians each strike one note. Sound mixture is imagine an observer asked to make the attempt. a subjectively oriented concept. It refers to How can we evaluate the correctness of the what the listener hears. If one note is audible, observer's answers? Adopting replicability as while two others are struck simultaneously in Color Mixture 315 a soundproof enclosure, no chord has been out of which any spectroscopic apparatus such heard by that listener. as a prism manufactures the different mono- Visual phenomenology includes no such chromatic rays, by a process which is physi- thing as a simultaneous occurrence (mixture) of cally equivalent to the mathematical resolution two colors in which an observer sees only one of an arbitrary function into periodic terms by of the colors. A possible exception is that some Fourier's integral theorem" (Whittaker[ 1910] theories of color blindness imply that this 1951, 1:17). Fourier's integral theorem asserts might occur. These theories leave us that any periodic function of a single variable wondering whether a mixture of red and black p, which does not become infinite at any would look black to an individual with phase, can be expanded in the form of a series deficient sensibility to red. consisting of a constant term, together with a A mixture of colors, unlike a mixture of double series of terms, one set involving sounds, occurs only if two or more colors are cosines and the other sines of multiples at the seen at the same time and place. The require- phase. ment for simultaneity in space and time appears White light as a mixture of colored rays is to violate the principle that colors cannot a figure of speech. Color mixture is as elliptic- interpenetrate. Identifying a green object as a cal, though we can assess the issue without mixture of blue and yellow is inconsistent with Fourier's integral theorem. Mixing colors is the idea that nothing is both entirely blue and impossible, because colors cannot be isolated entirely yellow. Or nothing can be both blue from objects. Whether these objects are colored and not-blue. lights or colored pigments makes no difference. Mixture is a troublesome term in the expla- I cannot take blue chairs and put the blueness in nation of light in the physical sciences. For one pile and the chairs in another. Nor can I Jenkins and White, the theories of Gouy and accomplish this separation with blue pigments or others raise "the question as to whether New- blue lights. ton's experiments on refraction by prisms, Any statement about color mixture involves which are usually said to prove the compos- an ellipsis. If yellow and blue make green, this ite nature of white light, were of much sig- can only be verified by demonstrating that nificance in this respect . . . the view that the when yellow pigment is mixed with blue colors are manufactured by the prism, which pigment the pigment mixture looks green. was held by Newton's predecessors, may be Pigments, rather than "colors," have been mixed; regarded as equally correct" ( Jenkins and and pigment particles, rather than colors, have White 1957, 223). intermingled. For Whittaker, though Newton's prism Green need not be thought of as a mixture experiments demonstrated that "ordinary of blue and yellow colors. It is more ade- white light is really a mixture of rays of every quately identified as the color presented by a variety of color .... The word mixture must mixture of blue and yellow pigments. Or it is not be taken to imply that the rays of differ- the color of a mixture of those pigments in ent colours, when compounded together, pre- some cases. If the pigments consist of large serve their separate existence and identity chunks, each half an inch in diameter, a mixture unaltered within the compound, like two con- does not look green. It looks like a mixture of stituents in a mechanical mixture. On the con- pigment chunks in which some are blue and trary, as was shown by Gouy in 1886, natural some yellow. white light is to be pictured, in the undulatory For mixtures of light, we do well to follow representation, as a succession of short pulses, Whittaker with appropriate modification. 316 Color Mixture

White is not obtained in light by a mixture of (monochromatic) green spectral light. Blue + colors. Instead, white is the color of what "is to yellow = green is not a reversible equation. be pictured, in the undulatory representation, as Green = blue + yellow is not necessarily true. a succession of short pulses, out of which any Because of the ellipsis in any assertion that spectroscopic apparatus such as a prism colors are mixed, any color name possesses a manufactures the different monochromatic rays, double referend. Green properly identifies by a process which is physically equivalent to materials that are inherently green. It can also the mathematical resolution of an arbitrary apply to mixtures of blue and yellow materials, function into periodic terms by Fourier's integral though only if they look green. A jar of blue theorem." marbles and yellow marbles is not a jar of green Statements about mixing colors refer to the marbles. Although the familiar recipes of color color of a mixture of something other than theory suggest otherwise, green cannot be color: of light, pigment, or other substance reduced to blue + yellow. Nor is blue + yellow located in the three-dimensional world. Colors necessarily green. are indeed mutually exclusive and cannot interpenetrate. A green object is singularly green. Its greenness may or may not have been Using the Color Wheel to Predict created by mixing blue and yellow pigments. Color is located in the two-dimensional visual Color mixing is a misnomer, not a breach of the realm that we, as three-dimensional bodies, are visual limit that I cannot see blue at the same unable to enter physically. We are limited to time and place I see yellow. looking into it or at it, unable to mix colors because colors cannot be reached or touched. I cannot will piles of red and blue pigments to Mixed Colors Without Mixture intermingle as a single pile of powders. Nor can A compelling reason exists for identifying green I will myself to perceive the respective redness as the color of some mixtures, rather than a and blueness of two color spots combined into a mixture in its own right. Many greens are not single mixed color. We live in one world while mixtures in any understandable sense. Green looking into another, or the human condition is grass is not a mixture of blue and yellow grass. indistinguishable from that model. The green chlorophyll that gives grass its color What is nominally called mixing colors is is not a mixture of blue chlorophyll and yellow an action undertaken in the three-dimensional chlorophyll. Many green substances, including world, a combining of pigments, lights, or green pigments, are inherently green, therefore other colored materials. If the term is under- not open to explanation as mixtures in any stood in this limited sense, any color can be sense. mixed with any other. More exactly, any two Examples of traditional green pigments colored materials can be mixed, unless they used by artists include chrome green, Hooker's repel or destroy one another. The question of green, and viridian. The colors of some of how many ways colors can be mixed reduces these green pigments (such as cadmium green) to that of how colored materials can be are difficult or impossible to replicate by mix- combined. ture of blue and yellow pigments. Newton Maxwell identified seven ways of mixing found a comparable phenomenon in light. colors (Maxwell [1890] 1965, 126-54). I have Mixing blue and yellow spectral rays, he added eight others to his list. Among the fif- produced a ray of green. This, however, was teen, only two are ordinarily considered in not the same green as a ray of unmixed explanations of color mixture. We are told Color Mixture 317

what to expect from mixture of colored lights Figure 34-2. Results of mixing light. The results of mixing lights of different colors are rarely similar to and from mixture of colored pigments. the obtained by mixing pigments or by mixing The traditional narrowness of focus suggests colors Maxwell discs. (Adapted from Helmholtz [ that pigments and lights have been regarded as 1909] 1962:129.) paradigmatic, as symbols of the material and 1. red + violet = purple immaterial aspects of the threedimensional 2. red + indigo-blue = dark pink 3. red + cyan-blue = pale pink world. But it was also a genuine discovery, 4. red + blue-green = white surprising when first made, that colored lights 5. red + green = pale yellow 6. red + green-yellow = golden yellow and colored pigments rarely behave similarly in 7. red + yellow = orange mixtures. White light can be synthesized from 8. orange + violet = dark pink 9. orange + indigo-blue = pale pink rays of the spectral colors, or it can be 10. orange + cyan-blue = white separated into them. As was known long before 11. orange + blue-green = pale yellow 12. orange + green = yellow Newton, these accomplishments cannot be 13. orange + green-yellow = yellow imitated with paints. 14. orange + yellow = (?) 15. yellow + violet = pale pink Color mixing can involve both lights and 16. yellow + indigo-blue = white pigments. A blue light might be allowed to fall 17. yellow + cyan-blue = pale green 18. yellow + blue-green = pale green on a yellow object. The object, under that set of 19. yellow + green = green-yellow circumstances, looks neither blue nor yellow 20. yellow + green-yellow = (?) 21. green-yellow + violet = white but some combination of the two. The blending 22. green-yellow + indigo-blue = pale green of colors on spinning Maxwell discs, which 23. green-yellow + cyan-blue = pale green 24. green-yellow + blue-green = green Maxwell pronounced equivalent to the blending 25. green-yellow + green = (?) of rays of colored light, eludes classification as 26. green + violet = pale blue 27. green + indigo-blue = water blue either pigment mixture, light mixture, or a 28. green + cyan-blue = blue-green combination of the two. 29. green + blue-green = (?) 30. blue-green + violet = water blue The chart of results when mixing colored 31. blue-green + indigo-blue = water blue lights in figure 34-2 is adapted from Helmholtz, 32. blue-green + cyan-blue = (?) 33. cyan-blue + violet = indigo-blue who warned that "mixed pigment does not 34. cyan-blue + indigo-blue = (?) give at all a colour that would be the resultant of mixing the two kinds of lights that are reflected separately from each of the ingre- dients" (Helmholtz [1909] 1962, 2:122). Helm- a historical oddity: the invention of the wheel is holtz's chart, indicates that red and cyan blue attributed to Newton, who used it to show his lights mix to form light that is pale pink. ideas about light. To Newton's surprise, the Orange and green mix to form yellow. Green- spectrum that fell on his wall was not circular, yellow and violet produce white. Red and but oblong, a narrow rectangle, with a width green make pale yellow. Golden yellow is a approximately five times its length. The oblong result of mixing red and green-yellow. Even familiarly had red at one extreme, violet at the allowing for the ambiguity of color names, other. these results are distinctive, different from Newton's rearranging of the spectral band those obtained with pigments or Maxwell into circular form became accepted practice discs. because it provided a more complete account The greater number of color wheels in print of visual experience. A continuum of reddish are designed to teach color-mixing principles violet colors lies "between" red and violet. to art students. They show what can be This range is absent from the spectrum or rain- expected in mixing paints ( pigments). This is bow, which does not include the full gamut 318 Color Mixture

of hue. To explain the discrepancy, the argu- ticular bright yellow pigment, a bright orange is ment is sometimes offered that the solar spec- obtained. Rarely is a mixed orange dull if trum is really circular as it emerges from a mixed from a bright red/yellow pair. prism, although it looks linear. For Gordon A smaller proportion of instances exist in Lynn Walls, "the spectrum really has no which bright blue pigment can be mixed with ends-it only seems to have, due to the way bright yellow pigment to produce a genuinely in which a prism forms it. Really it is a closed bright green. Sometimes the green is dull or entity, for red and violet are adjacent, drab. It deviates from any variety that would be psychologically-their mixture results in pur- widely identified as "a really good green." ple, which lies outside the spectrum but fills Among available red and blue pigments, a the gap between red and violet in a spectrum small proportion of instances exists in which which we might imagine bent into a ring" mixture of the colors in some red/blue set will ( Jacobson 1948, 116). provide a "good" (bright) purple. The dull We too often accuse our eyes of mislead- purples often obtained from mixtures of red ing us, and Walls's proposition is unconvinc- and blue have a grayish look. Although just ing. In visual experience, a distinction cannot two colors are involved in a red/blue mixture, be made between the ways color phenomena the grayish cast recalls the tendency of tertiary really are and the way they (really) look. The (three-color) mixtures with a high blue or color wheel, Newton's mangled solar spec- purple (blue/red) component to look more like trum, remains popular because reliable, to a grays than like browns. Blue has an affinity degree, in is predictions about color mixtures. with black and gray. Children can be taught such formulas as "blue Orange, the only secondary color mixed mixed with yellow makes green." without blue, comes closest to following the Because of its idealized nature, the wheel prediction of the color wheel that primary colors communicates ideas about color mixture that ought to mix to form secondary colors no less are wrong along with some that are roughly bright than they are. Mixed greens and purples, right. Two bright colors, when mixed, may not which require blue, deviate from this standard. produce a third that is similarly bright. Among Why is it easier to mix bright oranges than artists' pigments, phthalocyanine blue and bright purples? Although the phenomenon is lemon yellow form a bright green. Ultramarine rarely recognized or discussed, I think we can blue and cadmium yellow medium, an equally find an answer. bright pair, form a duller green with an olive The English language encourages an cast. Some bright blues and yellows yield a emphasis on hue when identifying colors, a product color approaching the outer limits of stress that downplays the value component of what anyone might want to call green. the primary colors. Spectral yellow is lightest, Blue and yellow make green when mixed, closest to white. Spectral blue is darkest, closest but only in the limited sense that they will not to black. Imagine ranking red, yellow, and blue produce pink or orange. The predictions made by assigning numbers that grow higher as the by the color wheel are approximate. They color grows darker: appear to have greater applicability for second- yellow 0 dary colors near the red end of the spectrum red 1 (orange), less for those near the violet or pur- blue 2 ple. Among available red and yellow pigments, many pairs exist in which, when that particu- We can next assign to each secondary color lar bright red pigment is mixed with that par a numerical value that is the average of the Color Mixture 319 numbers assigned to the two primaries mixed to present in all modern paints. Some, including obtain it. fluorescent paints, are made with liquid dyes. The process of making paints from dyes can orange 0.5 include a secondary vehicle that behaves as a green 1.0 pseudopigment though not inherently colored purple 1.5 itself. M. G. Martindill, a pigments and coating For the secondary colors, the higher the consultant, described a method by which number, the more chance that mixing bright "synthetic fluorescent dyes can be dissolved at primaries will lead to a dull, grayish, or low concentrations in a transparent liquid, darkened version of that secondary. The value usually a resin, and the mass than solidified to component of the primary colors evidently form a solid solution which can be ground to plays a role in color mixing that reflects some form a pigment" (Martindill 1988, 188). physical or optical law. Modern manufacturers also add powders that affect the color of the paint, yet are neither carriers of coloring material nor pigments in a Methods of Mixing Pigments conventional sense. In modern alkyd and Among the seven methods of color mixing polyurethane house paints and varnishes, a listed by Maxwell, the most familiar is the transparent, light-diffusing powder sinks to the mechanical mixture of colored powders. These bottom of the can when the paint stands. This are usually in the form of pigments. Although a powder accounts for the difference between flat mixture of household cleanser and face powder and matte varnish; between gloss black and flat has little value as a pigment, it illustrates, as black; between shiny paint and the same color easily as can be shown by pigments, the with a nonglossy surface. mechanics by which dry powders enter one Many characteristics of paints have more to another's interstices. do with the vehicle than with the pigment. Egg The traditional method for making paint tempera looks paler than oil paint and dries was invented in the Paleolithic era and is as old faster, although the same pigments are used in as the art of painting. It was used in the both. Water-base paints look darker when wet, Altamira and Lascaux caves. Particles of complicating the problem of matching colors. colored material (pigment) are suspended in a With house paints, which are not intended to paste or liquid vehicle. Without a vehicle to last as long as artist's paints, matching is bind them, powdered pigments cling to a difficult if a color has changed by exposure to surface with difficulty, a limit familiar to artists light or air. Repainting woodwork is preferred, who make drawings with chalk or pastel. The for this reason, to touching up spots where old paints applied to the walls of Paleolithic caves paint has chipped. appear to have been made from dried and pulverized clays, as well as other powders. These powders were probably mixed with Mixing Lights animal fats. Vehicles popular at later times Maxwell identified four methods of mixing include egg yolk, gum arabic, wax, oil, varnish, lights without regard to the source of the and acrylic medium. lights. These include superposition of beams Because a vehicle is present, mixing paints of different colors on an opaque screen, unit- usually involves combining colored liquids, ing beams by passing them through a prism, pastes, or gels, a process uneasily characterized and uniting two or more beams through a dou- as mixing pigments. Pigment powders are not bly refracting prism. In a fourth method, origi- 320 Color Mixture

nated by Helmholtz, beams are united by a closer to that suggested by Maxwell, No Kohler transparent surface that reflects one and devised goggles with lenses tinted blue on the transmits the other. In this method two colored left half and yellow on the right half (Kohler wafers lie on a table with a glass plate between 1972, 111). Observers saw through blue glass them. The glass is placed so that the reflection when looking to the left, through yellow when of one colored wafer corresponds with the looking to the right. Subjects adapted to these image of the other seen through the glass. devices. They learned to ignore the bisected Because a reflection on glass is faint, the color field presented by Kohler's goggles, much technique is similar to looking at a colored as wearers of bifocals learn to ignore a field of object through the of another object. vision split horizontally. Because the eye, or the mind, edits without conscious intervention, ease of adaptation is a Optical Mixture common phenomenon. For example, many Two methods of mixing color offered by years ago, I noticed that straight lines looked Maxwell play on the limits of acuity of the curved. The condition was diagnosed as human eye. In one, different colors can be astigmatism and since then has been corrected successively presented to the retina, as on by eyeglasses. I often wonder what purpose is spinning Maxwell discs. In the other, a different served by the correction. Straight lines soon color is presented to each eye by, say, asking a ceased looking curved whether or not I wore my subject to wear eye glasses with one red lens eyeglasses. I could not make the lines look and one green lens. In either of these methods, curved when I tried. Yet the astigmatism, a fault which cannot be explained solely in terms of in curvature of the lens of the eye, has not gone laws of color mixing, any mixture occurs in the away and is regularly discovered in eye subject's eye. examinations. Maxwell discs, and Maxwell's scientific Many parallel mechanisms exist. The nor- ideas transmitted through the writings of mal eye ignores the blind spot that exists at its Ogden Rood, inspired the French Impres- center and has no conscious awareness of the sionist painters to their thoughts on broken overlap of the visual fields for right and left eye. color, small spots of color meant to mix Elderly people rarely complain that everything optically in the eye of the viewer. In Maxwell's looks yellow, though the lens of the eye second method, the red and green eyeglasses, becomes yellowish with age. Kohler contended mixture may not occur in any ordinary sense. the eye's ability to adapt can be overloaded, a The respective perceptions of redness and plausible supposition. The condition called greenness would be directed to different sides of eyestrain is effectively an overload of the eye's the brain, which might accept the reading of a ability to adapt to difficult conditions. The issue, dominant side. The result would be perception in adaptation of the eye, is that a one-to-one of either red or green, not both at once or any correspondence does not always exist between third color that could be called a mixture of the what we see and what is available to be seen. red and green. We need to know more about how and what the Experiments with eyeglasses and goggles eye edits, and under what circumstances. have been undertaken by G. M. Stratton, The- odor Erismann, and James J. Gibson, using col- ored lenses and prisms that modified curvatures, created rainbow fringes around Other Methods objects, turned the visual field upside down, I have added the following methods for mix- and created other distortions. In an experiment ing colors to Maxwell's list, though the com- Dolor Mixture 321 pilation is still not exhaustive. bung. In glazing, a layer of oil paint, thinned A colored object or pigment can be to transparency or semitransparency by the illuminated by a colored light, as when a blue use of varnish, is superimposed on an opaque light shines on a yellow table. The technique is underlayer of another color. The technique is used in theatrical lighting and cinematography. used in Renaissance and Baroque painting, In the black light shows popular in the 1960s, sometimes over a monochrome underpainting the illuminating light was ultraviolet. in which white is mixed with the olive green A colored object can be viewed through pigment called terre vert (green earth). As in colored liquid, as when a lemon is immersed in the opalescent skin tones in Rubens's paintings, a glass tank of blue copper sulfate solution. the color of the underpainting is visible through, Two colored liquids can be mixed, as in but modified by, the layers of glaze. In combining green ink with blue ink. scumbling, a thin layer of fairly opaque paint Colored objects can be looked at through does not completely veil, and blends with, the anything transparent that is colored: colored color over which it is applied. The oncepopular glass, colored liquid, colored gelatin filters. technique of tinting black-and-white Practical application of the technique can be photographs simulates the uses of glazes in oil seen in sunglasses, stained-glass windows, and painting. the filters used by photographers. The squid, in Colors can be mixed by using movie film a variation, squirts black ink to destroy the and a projector. Film can be projected in which transparency of water and wraps itself in an blue frames alternate with yellow ones. This artificial night. would have to be done at higher speed than the Small dots of one color can be placed normal twenty-four frames per second, which among small dots of another until the surface is create an impression of blue flashes alternating covered. In art, the device appears in the with yellow flashes. broken color of the Impressionist painters, Chemical substances, as the alchemists refined to the smaller color spots of the knew, impart characteristic colors to flames. pointillist technique used by Seurat and the The phenomenon is the basis of spectroscopy, neoImpressionists. Modern three-color and which enables modern astronomers to study fourcolor process printing, used to reproduce the colors of the light of distant stars. The gas color illustrations in books and magazines, is a in kitchen stoves burns with a blue flame, more recent adaptation. although wood fires tend more toward red, Fabric weaving, discussed by Chevreul, orange, and yellow. Copper and its salts turn provides a related medium for color mixing. flames green, and so forth. A fine mixture of In a fabric woven with red woof and yellow powders, each imparting a different color, warp, the small spots of color fuse to a single should give flames of several colors or a mixed color if looked at from a sufficient distance or color when burned. if the yarns are very fine. In nature, a distant multicolored object, say, a building of bricks of several colors, fuses to a single color. We Teaching About Color Mixture cannot anticipate the color the building will One argument in favor of the limited type of look from six blocks away if 30 percent of its color theory diagrammed in color wheels is bricks are red, 12 percent are tan, and 58 per- that learning about primary, secondary, and cent are yellow. The color wheel is not suc- tertiary relationships imparts understanding of cessful in predicting results of optical mixture. how to mix colors. To the extent that art stu- In traditional oil painting techniques, mixed dents and others are to be educated about mix- colors can be produced by glazing and scum- ing paints, what ought to be taught is largely 322 Color Mixture

inconsistent with what the color wheel implies. dure of choice, more easily controlled, is to Consider the proposition that red paint, yellow make fine adjustments to the available green paint, and blue paint can be mixed to form paint closest to the required color. This paint brown paint. Does anyone do this? Should the can be lightened, darkened, made more yellow, procedure be recommended? modified as desired. Anyone who asks how to Commercially available brown paint, mix red paint ought to be told, as the color including that intended for artists, is not made wheel implies, to buy a tube of red paint. The from a mixture of red, yellow, and blue pig- person also should be reminded, though this is ments. The pigment, or coloring matter, used in not clear from color wheels, that if two its manufacture is inherently brown. No different reds are on hand, a variety of others incentive exists for a manufacturer, or anyone can be made by mixing them. else, to turn to alternate procedures. Pigments We are not advised to manufacture our own that approach the spectral hues in color are water just to prove it can really be syn- costly. Brown pigments are inexpensive in thesized from hydrogen and oxygen. Know- most cases. A large and important group of ing which colors mix to form brown has a traditional brown pigments, the earth colors negative value as an admonition to stay away (for example, umber, sienna, ocher) are the from these mixtures. Pissarro's advice not to least expensive of all pigments. As the name mix colors is meant to remind that mingling suggests, these pigments are made from earth, bright colors from different hue ranges results which varies in color in different locales in a brown or tertiary, which may not be the because of differences in composition. end desired. The best advice for an art student confused Mixing purple is as ill-considered an enough to contemplate mixing brown from, endeavor as mixing brown, though for other say, cadmium red, cadmium yellow, and reasons. Inherently purple pigments are avail- ultramarine blue is to abandon the project. able and are brighter than purples mixed from Buying brown paint is more cost-effective. The the more commonly available reds and blues. narrower assertion that blue and yellow make At the same time, an interesting range of pur- green is equally limited in scope. The equation ples and violets can be created by mixing rose is not true if reversed. Green is not necessarily madder with any black, though color theory composed of blue and yellow in the strict sense allows for this range only when red is mixed that H20 must include hydrogen and oxygen. with blue. Limiting its pedagogical usefulness, Green paint is usually manufactured from the color wheel provides answers that are too inherently green pigments, not those that are short, too simplistic, too certain, as if all the blue and yellow. answers were known. It was never intended for A person who wants to paint a chair in a the purpose of instilling in students the green of his or her choosing is ill-advised to confidence to ask questions about color and to begin with yellow and blue paints. The proce- reason about what can be observed. CHAPTER 35 Additive/Subtractive Theory

The theory of the additive and subtractive varieties of color mixing still stands uncontested today. Maurice Grosser, The Painter's Eye

light, as Newton showed, can must be because paint is less pure than light. I be produced by recombining the do not know what pure means in the context, or spectral rays derived (or manufac- why paint ought to be regarded as impure tured) from it. The effect cannot be replicated because different from light. How is the with paints, where no mixture of colors yields explanation to be extended to Maxwell discs? white. Not can it be achieved with Maxwell The discs are as naughty, or as impure, as discs, or by any form of mixture that involves pigments. They too refuse to behave as we both lights and pigments. Newton's discovery think they ought to. will not support the broad generalization drawn Those seeking a more satisfying technical from it. White paint, unlike white light, is not a explanation can turn to what is loosely known mixture of colored components. The as the additive/subtractive theory, familiar to discrepancy is inconvenient because it under- most people who have studied art. The general mines the proposition, in popular belief and the form of the theory varies among proponents, physical sciences, that color can be understood but it begins with the proposition that rules in terms of light. regulating colored lights have an inverse rela- To smooth over the difficulty, the thought tionship to those governing colored pigments. is sometimes offered that if white paint can- When lights of two colors are combined, not be made by mixing colored paints, this the theory asserts, their wavelengths ( ex-

323 324 Additive/Subtractive Theory

pressed in millimicrons) are added together. Figure 35-1. Wavelengths of the spectral colors. The wavelengths of the third color obtained Wavelengths of the spectral colors are given variously by various authors. But the range is from the mixture is equal to the sum of the generally from approximately 380 to 780 wavelengths of the two constituent colors. This millimicrons. (After A. C. S. Van Heel and C. H. F. charming nonsense compels an untenable Velzel, What is Light? [ New York: McGraw Hill, conclusion. It implies that mixing red lights of Spectral Color Wavelengths two different wavelengths-say, 645 millimicrons and 650 millimicrons-would yield light with a wavelength of 1295 millimicrons, As figure 35-1 indicates, this is beyond the upper extreme of the visible light sector of the electromagnetic scale. In everyday language, the fallacious prediction is that mixing two slightly different shades of red light would yield wavelengths. In everyday language the a third color human beings could not see. It fallacious prediction is that mixing two slightly would lie in the infrared range. If we mix, say, different shades of red pigment, no less than seven or eight colors, and have more mixing two slightly different shades of red wavelengths to add together, the sum can carry light, would yield a third color human beings us beyond infrared to the ranges for radar and were unable to see. The theory is exceedingly radio waves. strange. When colored pigments are mixed together, In books dealing with color and visual the theory doggedly continues, the wavelength perception, the additive/subtractive theory is (in millimicrons) of one color is to be subtracted alluded to more often than explained, and rarely from the wavelength (in millimicrons) of the examined in embarrassing detail. Like many other. The wavelength of the third color strange propositions in color theory, its obtained from the mixture is equal to the foundation is said to lie in physics, with no difference between the wavelengths of the two exact pinpointing of provenance to be reached. constituent colors. Presumably the smaller Bertram Cholet skims over the theory lightly, as wavelength is to be subtracted from the larger, does the psychologist R. L. Gregory (Cholet to avoid negative wavelengths. The theory n.d., 10-14; Gregory 1966, 119). Recent books assumes that any mixed color is limited to no mention no particular problem with it and, if more than two components, which need not be they ignore it, offer no new theory. References the case. If a mixed color has three, eleven, or to adding and subtracting light waves and to nineteen components, how can anyone additive and subtractive primaries (the primaries determine which wavelengths to subtract from in pigment and light) are still common. which? Additive/subtractive theory hazards no One of the more painfully detailed opinion. explications is found in Light and Vision, a Limiting the task to just two colors, volume prepared in the late 1960s by the subtracting wavelengths is as dubious as adding science editors of Life magazine. The Life them. We are to understand that mixing, say, editors introduce an additional embroidery in red pigments of two slightly different their explanation of the discrepancy outlined: wavelengths (645 and 650 millimicrons) would yield pigment of a color corresponding to 5 millimicrons. This lies beyond the lower The answer to the riddle [ of why white extreme of the electromagnetic scale for visible paint cannot be mixed from colored Additive/Subtractive Theory 325

paints ] lies primarily in the totally no mechanism for adjusting the calculations to different ways colors are achieved with take account of different proportions. Red and light and pigments. The differences are yellow make orange, but only in the sense that analogous to addition and subtraction in orange is a range of colors rather than a single mathematics. All spectral colors can be shade. A continuum of variations, ranging from created by adding, in varying degrees of reddish orange through yellowish orange, can intensity, different amounts of three be created from any given red and yellow by primary components of light. Pigment varying proportions. colors, on the other hand, are arrived at A very yellowish orange and a very reddish by subtraction . . . . Green is relatively orange cannot be expected to have the same short [in wavelength], about 500 wavelength, though each can be mixed from the millimicrons; red is extremely long, about same red and yellow. If each shade of orange is 700 millimicrons; red is extremely long, assumed to have its own wavelength, about 700 millimicrons. The eye averages additive/subtractive computations must lead to a these two wavelengths and sees one of range of wavelengths, not a single averaged 600 millimicrons, which is in the yellow wavelength. Any theory that predicts the result sector of the spectrum. (Mueller and of color mixtures by computations involving Rudolph 1966, 98). wavelength must lead to more than one numerical answer in order to take account of Because the explainers change premises in proportion. midparagraph, whether the wavelengths of two Other absurdities in additive/subtractive colors of light should be added or averaged to prediction derive from its failure to take arrive at the wavelength of the mixed color account of the circular relationship among the remains unclear. The question is whether either hues, of the two continua between red and adding or averaging makes sense. violet. Mixtures of red and violet paint yield Black and white pigments mix to form gray, red-violet paints, which cannot reflect aver- a phenomenon not to be explained by any aged wavelengths. The range of red-violet adding, subtracting, or averaging of colors is not included in the electromagnetic wavelengths. The achromatic colors are without scale and is consequently devoid of assigned assigned wavelengths, other than that black is wavelengths. Furthermore, additive/subtrac- loosely called "no" light (no wavelengths), and tive theory, as understood by the science editors white "all" wavelengths. Gray is presumably on of Life, incorrectly predicts that red and violet a continuum between. Because all has no will mix to form yellow (580 millimicrons), numerical equivalent (none might be which has a wavelength representing the understood to mean 0), no computations can be average of the respective wavelengths of red performed. Thus, additive/subtractive theory (780 millimicrons) and violet (380 mil- provides no predictions about mixtures limicrons). involving colors that lack assigned wavelengths, Among other nonsensical predictions from that are not included on the electromagnetic taking the average of wavelengths, green (525 scale. The excluded colors consist of the millimicrons) is incorrectly predicted as the achromatics and all other nonspectrals, in result of mixing violet and orange (450/600). aggregate a larger class than the spectral colors. Blue-green (500) is incorrectly predicted as the The additive/subtractive theory is opera- result of mixing violet and yellow (400/600). tionally impossible to apply. It never specifies In both of these examples, adding or subtract- proportions for the colors to be used and has ing yields a result as absurd as averaging. 326 Additive/Subtractive Theory

If wavelengths ought to be added for light, not be made by mixing A and B. Nor can its red light (761 millimicrons) is incorrectly wavelength be obtained by adding, subtracting, predicted as the result of mixing two violet or averaging the wavelengths of any other two lights (380 and 381 millimicrons). If wave- colors. lengths ought to be subtracted for pigments, bluish violet (400), rather than reddish violet, A + B ≠ X is incorrectly predicted as the result of mix- A - B ≠ X ing red pigment (780 millimicrons) with vio- (A + B)/2 ≠ X let pigment (380 millimicrons). Mixing orange The two sets of formulas are not consistent. and blue (630/450) or green and yellow Term X represents wavelength (in millimicrons) (550/570) is incorrectly predicted to yield of a primary color. No number X exists that results completely off the scale, whether one cannot be expressed as the sum of two other wavelength is added to or subtracted from the numbers, the difference between two other other. numbers, or the average of two other numbers. The faulty predictions of additive/subtractive Either the result of color mixture cannot be theory derive from its questionable logic, too predicted by arithmetical computations on the eccentric to patch or repair. Assume that any wavelengths of the colors or primary colors do color can be correlated with a numerical not exist. A third possibility is that notation indicating its wavelength in additive/subtractive theory is wrong and no millimicrons. X is the numerical value of the primary colors exist. wavelength of a primary color. A and B are the Mathematical sophistication is not required numerical values of the wavelengths of two to understand why the predictions are absurd colors that may or may not be primary. C is the when wavelengths are added or subtracted. The wavelength of the color that results from mixing numerical value for the wavelength of mixed A and B. These objects exist in a common color C must lie between 380 and 780 medium, whether pigment or light. (millimicrons), the visible light sector of the The additive/subtractive theory predicts, electromagnetic scale. But the numerical values according to its major versions, one of three for A and B lie in the range from 380 to 780. results when A and B are mixed. (1) The Among solutions for the formula A + B = C, numerical values of A and B will be added to where A and B are each more than 380 but less provide the wavelength of product color C. (2) than 780, over 90 percent of the value for C The numerical values of A and B will be exceed 780. subtracted, one from the other, to provide the The problem with subtraction is similar. wavelength of color C. Or, (3) the numerical The value for C cannot be less than 380 if it values will be averaged, again providing the is to correspond with a wavelength in the visi- wavelength of product color C. The possibilities ble light sector of the electromagnetic scale. can be notated as follows. Among solutions for A - B = C, where A and B are each more than 380 but less than 780, a A + B = C = additive mixture (light) large number of values for C will be less than A - B = C = subtractive mixture (paint) 380. (A + B) / 2 = C = averaging (light or paint) The main problem that intrudes in average- Primary color X, by definition, cannot be ing is different. In physics, the relationship made by any type of mixture of any other among the hues is diagrammed in a linear man- colors. Certain negative formulas can be given. ner. The wavelength in millimicrons of yellow They state that color X, because primary, can- (580) is an average between the extremes of Additive/Subtractive Theory 327 red and violet, 780 and 380. This solves the studying the results of the proposed equation, but gives a wrong answer. We do not computations. The starting point was probably obtain yellow by mixing red and violet. the observation that white light can be explained as a mixture of colored rays. It was (A + B) / 2 = C (780 + 380) / 2 = 580 assumed that white paint ought to be open to a similar explanation, an assumption for which no Two continua exist between red and vio- basis exists. Mixing pigments of every hue let, because the relationship among the hues yields brown, gray, black, or near black. is circular, not linear. The color wheel, though Because pigment mixture and light mixture it has shortcomings, diagrams this in a correct could not be shown to be similar, they were manner. Red-violet is the color obtained by assumed to be opposite. White is popularly mixing red and violet. It lies on the second called the opposite of black. continuum, the continuum absent from the A probable reason for concluding tha line diagram used in electromagnetic theory. wavelengths of light had to be added is tha the The formula for averaging (A + B)/2 = C, result of mixing all the colors of light wa would have to be expanded to provide for two supposed to be white, loosely identified as a: answers, sufficient to indicate which con- wavelengths, an infinity of wavelengths. . tinuum we ought to look on to locate a given probable reason for concluding tha color C lying halfway between A and B. Either wavelengths of pigments had to be subtracte is (A + B) / 2 = C1 (the color on continuum 1), that the result of mixing every color of pair or (A + B) / 2 = C2 (the color on continuum was supposed to be black (loosely identifie as 2). a nullity, or no wavelengths). The theor was The theory cannot be patched to accom- programmed for absurdity. A hodgepodg of plish this end. The circular nature of hue unexamined popular misconceptions abot relationships cannot be accommodated in color was accepted at face value. They wer electromagnetic theory. No range of assumed to have a rational or scientific four wavelengths exists for reddish violet colors, no dation, and a scientific explanation was dil second continuum. The electromagnetic scale, a gently constructed. linear scale, cannot be adjusted to provide one. Less diligence was applied to examining the Nor does any way exist to compute an results of the proposed computations or the arithmetical average that would allow for two problems at hand. Three difficulties are insur- different results. mountable. First, mixing two colors never Another arithmetical limit is that wavelength yields a single product color, one reason range for violet (450 to 380) is less than that for additive/subtractive theory can never work. either blue (500 to 450) or red (780 to 630). Any two colors can be mixed in various Although violet is the result of mixing red and proportions and yield a continuum or range blue, no average of wavelengths in those ranges of product colors. Second, all colors do not can ever fall in the violet range. If (A + BY2 = have assigned wavelengths. The electromag- C, C cannot be less than A/2 + B/2. All values netic scale includes only the hues and their for C, in this case, fall considerably below that intermediaries with the exception of the red- limit. violet range. Third, two roads lead from red The author of additive/subtractive theory to violet, and we need to know which road is mercifully anonymous. The theory appar- to take, which direction to travel in the hue ently came into being when a chain of assump- circle. Orange, yellow, green, and blue are tions were made that were not tested by between red and violet, as on the electromag- 328 Additive/Subtractive Theory

netic scale. The range of red-violet colors is series turned to averaging as an aid to produc- also between red and violet, a phenomenon ing more satisfying computations. The proposal electromagnetic theory cannot accommodate. that the eye averages wavelengths demolishes In mixing red and violet paints, the result is any possibility of explaining in terms of red-violet. The color does not average out to wavelengths why mixtures of pigments yield yellow. As for the purported oppositeness of different results from mixtures of lights. It light mixture and paint mixture, Helmholtz implies that the eye perceives an average identified purple-not yellow-as the result of among wavelengths whether lights or pigments mixing red and violet lights (see figure 34-2). are involved. No computational opposite to When adding and subtracting of wave- averaging is available to shore up the proposal lengths is tested, the computations yield non- that mixtures of lights operate according to sense. This must have been apparent to the first opposite rules from those governing mixtures person who tried the computations. Somebody of pigments. likely assumed that a wrong form of Averaging has its own problems, doggedly computation was being used. The sensible overlooked. Additive/subtractive theory had question, not asked, is why we believe any- acquired a reputation for being scientific, tech- thing significant can be discovered about color nically unimpeachable, before it came to the mixture by arithmetic computations involving attention of the editors of the Life science wavelengths. series. The question, as often in color theory, is Sensing the futility of adding and subtract- how so ill-conceived an idea survived for so ing wavelengths, the editors of the Life science long. Notes

Chapter 1: Learning to Use Color Names ments,"J. Research NBS 31 (1943): 55; S. M. Newhall, 1. On ostensive definition, see Russell 1948, 63. On "Preliminary Report of the OSA Subcommittee on the ostensive learning of color names, see Rhees 1969. Spacing of the Munsell Colors," J. Opt. Soc. Am. 30 2. Webster's Second International Dictionary traces (1940): 617; S.M. Newhall, D. Nickerson, and D.B. Judd, yellow to the Old English geolu, which means yellow or "Final Report of the OSA Subcommittee on the Spacing of yellowish. Blue is from Middle English, Old High German, the Munsell Colors,"J. Opt. Soc. Am. 33 (1943): 385; D. and Old English words that appear to have had no meaning Nickerson, J.J. Tomaszewski, and T.F. Boyd, "Colorimetric other than bluish. White has no known root earlier than hwit Specifications of Munsell Repaints," J. Opt. Soc. Am. 43 (Old English), which means white. Black and gray are from (1953): 163. blaec and graeg (Old English), meaning black and gray. Brown from the Sanskrit bhru, apparently always meant brown. Chapter 5: Knowing How to Identify Color 1. See Geschwind 1969. For a more detailed report on this same patient, see Geschwind and Fusillo, Arch. Neurol. Chapter 3: Understanding Color Names 15 (1966): 66. For an earlier report on a case of this type, 1. For analysis and modifications of the Munsell sys- see H. Lissauer, Arch. Psychiat. Nervenkr. 21 (1889): 222. tem, see J.J. Glen and J.T. Killian, "Trichromatic Anal- ysis of the Munsell Book of Color," J. Opt. Soc. America 30 (1940): 609; W.C. Granville, Dorothy Nickerson, and Chapter 7: Light and Dark in Perspective C.E. Foss, "Trichromatic Specification for Intermedi- 1. For seventeenth-century publications on the spec- ate and Special Colors of the Munsell System,'' J. Opt. trum that predate Newton's report of his experiment, Soc. Am. 33 (1943): 376; K.L. Kelly, K.S. Gibson, and see Marcos Antonio de Dominis, De radiis visus et lucis D. Nickerson, "Tristimulus Specification of the Mun- in vitris perspectivis et hide tractatus (Venice, 1611); sell Book of Color from Spectrophotometric Measure- Rene Descartes, "Dioptrique" and " Meteores " ( Leiden,

329 330 Notes

1637); Marcus Marci, Thaumanthias, liber de arcu Chapter 10: Red-Violet, Blue, Brown, and Optical coelestri deque colorum apparendtun natura (Prague, Mixture 1648); F.M. Grimaldi, Physico-mathesis de lumine, 1. Bertrand Russell stated: "Physics is an empirical coloribus et iride (Bologna 1665); Robert Boyle, science, depending for its credibility upon relations to Experiments and Considerations Touching Colours our perceptive experiences (Russell 1948, 256). (London, 1664); Robert Hooke, Micrographica; or, Some Physiological Descriptions of Minute Bodies Made by Chapter 19: Complementarity in the Visual Field Magnifying Glasses (London, 1665). 1. For a description of the cinematographic tech- niques, see Raymond Fielding, The Technique of Spe- cial Effects Cinematography (New York: Hastings Chapter 8: Newton House, 1968), 259-321, which also includes a bibliog.- 1. Newton stated: "To the same degree of Refrangi- raphy of eighty-two articles dealing with the process. bility ever belongs the same colour, and to the same colour ever belongs the same degree of Refrangibility. Chapter 21: Hue, Color, and Culture The species of colour, and degree of Refrangibility 1. "A `red object' is an object which is otherwise proper to any particular sort of Rays, is not mutable by known than by the quality red; it is an object which has Refraction, nor by Reflection from natural bodies, nor been given a determined place in an order. The sensation by any other cause that I could yet observe. When any is an object which has not yet thus been placed. It is one sort of Rays hath been well parted from those of incorrect, then, to say that we can have sensations of other kinds, it hath afterwards obstinately retained its redness; redness is a concept; or to say that we have colour, notwithstanding my utmost endeavors to sensations of red. The sensation is of a red something, a change it" (Whittaker [1910] 1951, 1:14). red spot or area. And the discovery that the cause of the 2. "In the end it was shown that everything in physics sensation is a pathological irritation does not affect the can be explained either on the particle hypothesis or on objectivity of the sensation in the least. The red `that' the wave hypothesis. There is therefore no physical was there, and the fact that the object cannot be further difference between them, and either may be adopted in defined and verified does not make it any the less any problem as may suit our convenience. But whatever object" (Eliot 1964, G2). is adopted, it must be adhered to; we must not mix the two hypotheses in one calculation" (Russell 1948, 23). Chapter 22: Prime Minister Gladstone and the Blues See also: "Take, for example, the question of waves 1. See Wallace 1927, 5. See also H. Magnus, Die versus particles. Until recently it was thought that this geschichrliche Enrwicklung des Farbensinnes (Leipzig, was a substantial question: light must consist either of 1877); and H. Magnus, Untersuchungen fiber den waves or of little packets called photons. But at last it Farbensinn der Narurvolker (Jena, 1880). For other was found that the equations were the same if both authors on the use of color words in Homer and ancient matter and light consisted of particles, or if both authors, see: J. Soury, De 1'evolution historique du sens consisted of waves. Not only were all equations the des couleurs (1878); J. Lorz, Die Farbenbe2eichnungen same, but all the verifiable consequences were the same. bei Homer mfr Beriicksichrigung der Frage uber Either hypothesis, therefore, is equally legitimate, and Farbenblindheit (1882); A. de Keersmaecker, Les sens neither can be regarded as having a superior claim to des couleurs chez Homere (1883); Edmund Veckenstedt, truth. The reason is that the physical world can have the Geschichte der griechischen Farbenlehre; das same structure, and the same relation to experience, on Farbenunterscheidungsvormogen, die the one hypothesis as on the other" (Russell 1948, 256). Farbenbezeichnungen der griechischen Epiker, von Homer bis Quintus Smyraneus (1888); A. Clerke, Familiar Studies in Homer (1892), 294-302; N.P. Chapter 9: The Cause of Color and Light Benaky, Du sens chromatique Bans 1'anriquire sur le 1. For scattered experimental evidence that questions base des dernieres descouvertes de la prehistorie, de the classical correlation between colors and wave- I'etude des monuments ecrirs des anciens et des donnees lengths, see Land: "Color in images cannot be described de la glossologie (1897). in terms of wave length and, in so far as the color is changed by alteration of wave length, the change does not follow the rules of color-mixing theory" (Land Chapter 24: Tristimulus Theory and Metamerism [1959] 1961, 388). For a refutation of Land's views, see 1. See W.A.H. Rushton, "The Cone Pigments of the Judd (1979). For observations of the color blind that Human Fovea in Color Blind and Normal," Symposium suggest conclusions inconsonant with classical color on Visual Problems of Color (New York: Chemical Pub- theory, see C.H. Graham and Y. Hsia, "Some Visual lishing Company, 1961), 107. Rushton was challenged Functions of a Unilaterally Dichromatic Subject," in by Wald, who complained that although "Dr. Rushton Symposium on Visual Problems of Color (New York: had faithfully refrained from saying anything about a Chemical Publishing Company, 1961), 283-97. foveal blue-sensitive pigment it was nonetheless there." Notes 331

Chapter 25: Color and Form in Art of Colour: A Textbook for Students, Teachers of Art, 1. Stendhal, Article 7, journal de Paris, 9 October, and All Interested in Color (Leicester, 1935); Arthur B. 1824. Allen, Colour Harmony for Beginners (London, 1936); 2. Review of Durand-Ruel exhibition of Impressionist Arthur B. Allen, The Teaching of Colours in Schools paintings, by Pierre Wolff for Le Figaro. Reprinted in (London, 1937). Jean Renoir, Renoir, My Father (Boston, Little, Brown 3. Birren's books include Color in Vision (1928), & Co., 1962),158. Color Dimensions (1934), The Printer's Art of Color (1934), Functional Color (1937), Monument to Color (1938), The Story of Color (1941), Selling with Color (1945), and Color Therapy (1950), Chapter 26: Subjectivity and the Number of Colors 1. See Sargent: "Experiments show that, without New Horizons in Color (1955), Creative Color (1961), training, our eyes perceive easily about five degrees of Color for Interiors (1963), History of Color in Painting value (i.e., gray), beginning with white and ending with (1965), and Principles of Color (1969). He edited works black .... With a little training we can recognize and by Chevreul, Ostwald, Munsell, and Moses Harris, assign to their place in the scale about twice as many" sometimes heavily. An edition of Ostwald's Color (Sargent [1923] 1984, 62). See also Birren: "Psycholog- Primer is prefaced by Birren's editorial assurance that ical study has shown that the average person will read- the original text is "more or less intact" (Ostwald [1916] ily distinguish about nine steps from black to white" 1969, 17). Munsell's A Grammar of Color, though a (Birren, 1961, 24). See also Itten: "I have had many "magnificent volume," is pared to six pages, augmented interested and gifted students who were able to make by seventy-five contributed by Birren and others. The visible up to forty-four tone gradations between black explanation is that "Munsell was not much of a writer, and white" (Itten, 1963, 19). See also Chandler 1934, and though he could think and speak clearly and 69-70. See also Ostwald: "The number of distinguish- coherently, to all indications he found the task of author able steps of gray under normal conditions amounts to a difficult one. The best presentations of his views are by several hundred" (Ostwald [1916] 1969, 23). others" (Munsell [1921] 1969, 40). 2. See D. B. Judd, "Color Perceptions of Deutera- 4. The paintings Jacobson discussed are Giotto's The nopic and Protanopic Observers,"J. Res. Nat. Bur. Stan- Epiphany, Picasso's cubist Fruit Dish, El Greco's View dards (1948): 247-71. See also C.H. Graham and Y. of Toledo, Rousseau's La Cascade, Botticelli's Three Hsia, "Some Visual Functions of a Unilaterally Miracles of Saint Zenobius, Vermeer's A Girl Asleep, Dichromatic Subject," in Symposium on Visual van Gogh's Bedroom at Arles, Picasso's Red Tablecloth, Problems of Color (New York: Chemical Publishing Matisse's Harmony in Blue, Gauguin's Mahana No Atua, Company, 1961), 283-97. Renoir's Madame Charpentier and Her Children, and 3. See Wald 1961, 27. Cezanne's Still Life with Apples.

Chapter 30: Systematizers and Systems Chapter 31: The Logarithmic Gray Scale 1. Ostwald lived in the United States at the time, 1. See, for example, Cholet: "[Fechner's law of the although the university at which he was lecturing is threshold] indicates that between black, or the absence given differently by different authors. Bitten (Ostwald of light, and white, or the presence of total light, our [1916] 1969, 8) identifies it as M.I.T. Jacobson (1948, perceptions are stimulated by logarithmic progression. 1) lists it as Harvard, where Ostwald was visiting This means that for us to discriminate between black professor of physical chemistry and Ingersoll lecturer and a perceptible gray or the first presence of white on the immortality of man. added to the black requires a very small amount of 2. Ostwald's other works include Er and Ich (Leip- stimulus indeed. This is the absolute threshold. To dis- zig, 1936), Der Farbkorper (Grossbothen, 1919), Die criminate between this first gray and a second gray Harmonie Der Farben (Leipzig, 1918), Die Farbenfibel requires an amount increased by a small fraction of the (Leipzig, 1916), Der Harmonthek (Grossbothen, 1926), threshold amount of white light and to discriminate Levenslinien (Berlin: 1927), Letters to a Painter (trans. between the second gray and a third gray requires a A. W. Morse: New York, 1906), The Ostwald Colour further increase by the same fraction and the amount Album, (trans. J. Scott Taylor; London, 1932-35). For increases as a logarithm up to total white. From these textbooks based on the Ostwald system, see O.J. Tonks, studies a gray scale was organized by Wilhelm Ostwald Colour Practice in Schools: A Graded Course in Color and the grays were termed achromatic colors. Their Seeing and Using for Children Between the Ages ofFive reflectances plot at equal intervals on a logarithmic and Fifteen (London, 1934);J.A.V. Judson,A Handbook graph" (Choler 1953, 18). Bibliography

Albers, Josef. 1963. Interaction of Color. New Their Application to the Study of Nature and Haven: Press. Hints on the Composition of Pictures,&c.24th Alexander, C.1924.The Real Inner Secrets of Psy- ed.London:George Rowney and Company.Ltd. chology. Vol. 5 of Astral-Plane Phenomena. Barasch, Moshe. 1978. Light and Color in the Los Angeles: C. Alexander Publishing Italian Renaissance Theory of Art. New York: Company. New York University Press. Allen, Arthur B.1936.Colour Harmony for Begin Barnes, Albert C., and Violette de Mazia. 1933. ners. London: Frederick Warne&Company. The Art of Henri Matisse. New York: Charles 1937. The Teaching of Colours in Schools. Scribner's Sons. London: Frederick Warne & Company. Barr, Alfred H., Jr. 1951. Matisse: His Art and Allen, Grant. 1879. The Colour-Sense: Its Origin His Public. New York: Museum of Modern and Development. Boston: Houghton, Art. Osgood, and Company. Benaky,N.P.1897.Du sens chromatique dans I'an- Aristotle. Physics. Loeb Classical Library. tiquité sur le base des dernieres descouvertes de la . Metaphysics. Loeb Classical Library. prehistorie, de I'etude des monuments ecrits des Arnheim, Rudolph. 1956. Art and Visual anciens et des donnees de 12 glossologie. Perception. Berkeley: University of California Benson, William. [1868] 1930. Principles of the Press. 1966. Toward a Psychology of Art. Science of Colour Concisely Stated to Aid and Berkeley: University of California Press. Promote Their Useful Application in the Asher, Harry. 1961. Experiments in Seeing. Decorative Arts. London: Chapman & Hall. Greenwich, Conn.: Fawcett Publications. Beveridge, W. M. 1939. "Some Racial Differences Bacon, J.1925.The Theory of Colouring: Being an in Perception." British Journal of Psychology Analysis of the Principles of Contrast and 30:57-64. Harmony,In the Arrangement of Colours,With Bitten, Faber. 1949. "Functional Color in the

332 Bibliography 333

Schoolroom." Magazine ofArt 42 (April): 136-38. . land Publishing. Originally published as De la loi du 1961. Creative Color. New York: Van Nostrand contraste simultane des couleurs. Reinhold. Cholet, Bertram. 1953. Color Digest. New York: . 1969. Principles of Color. New York: Van Higgins Ink Company. Nostrand Reinhold Company. Clerke, A. 1892. Familiar Studies in Homer. . 1976. Color Perception in Art. New York: Van Cohen, I. Bernard. 1958. Issac Newton's Papers and Nostrand Reinhold Company. Letters on Natural Philosophy. Cambridge: Harvard Blaker, J. Warren. 1969. Optics. New York: Barnes and University Press. Noble. Cohen, Walter. 1957. "Spatial and Textural Boas, Franz. [1927] 1955. PrimitiveArt. New York: Characteristics of the Ganzfeld." American Journal Dover Publications. of Psychology 70:403-10. Reprinted, 1966, in Bosanquet, Bernard. [1892] 1957. A History Experiments in Visual Perception, edited by M. D. ofAesthetic. New York: Meridian Books. Vernon. Baltimore: Penguin Books. Bowie, Henry P. 1911. On the Laws of Japanese A Color Dictionary. 1931. New York: The North Painting. San Francisco: Paul Elder & Company. American Society of Arts, Inc. Boyle, Robert. [1664] 1964. Experiments and The Color Tree. n.d. New York: Interchemical Considerations Touching Colours: First Corporation. Occasionally Written. Among Some Other Essays, Crick, Francis. 1981. Life Itself. Its Origin and Nature. to a friend: and now suffer'd to come abroad as The New York: Simon and Schuster. Beginning Of An Experimental History Of Colours Croce, Benedetto. [1909] 1968. Aesthetic: As Science (1664). London: The Sources of Science. of Expression and General Linguistic. Translated by Bragdon, Claude. 1932. The Frozen Fountain. New Douglas Ainslie. New York: Noonday Press. York: Alfred A. Knopf. Cutler, Carl Gordon, and Stephen C. Pepper. 1923. British Colour Council. 1934. Dictionary of Colour Modern Color. Cambridge: Harvard University Standards. 2 vols. London: British Colour Council. Press. . 1938. Horticultural Colour Chart. London: Denis, Maurice. [1890] 1966. Definition of British Colour Council. NeoTraditionalism: From the Classicists to the Brown, Richard F. 1950. "Impressionist Technique: Impressionists. Edited by Elizabeth Gilmore Holt. Pissarro's Optical Mixture." Magazine of Art 43 New York: Doubleday, Anchor. (January): 12-15. Descartes, Rene. [1664] 1972. Treatise on Man. Bruno, Vincent J. 1977. Form and Color in Greek Translated by Thomas Steel Hall. Cambridge: Painting. New York: W. W. Norton Company. Harvard University Press. Burnet, John. 1950. Greek Philosophy: Thales to Plato. Dickie, George. 1962. "Is Psychology Relevant to London: Macmillan and Company, Ltd. Aesthetics?" Philosophical Review 71:285- 302. Burris-Meyer, Elizabeth. 1938. Historical Color Guide. Doerner, Max. [1921] 1934. TheMaterialsoftheArtist New York: William Helburn. and Their Use in Painting, with Notes on the Techniques of the Old Masters. Translated by Eugen Capra, Fritjof. 1975. The Tao of Physics: An Neuhaus. New York: Harcourt, Brace and Exploration of the Parallels between Modern Company. Physics and Eastern Mysticism. New York: Bantam Eastlake, Charles Lock, ed. 1970. [Goethe's] Theory of Books. Colours. Translated by C. L. Eastlake. Cambridge, Carnap, Rudolph. 1966. Philosophical Foundations of Massachusetts: MIT Press. Physics. Edited by Martin Gardner. New York: Eastman Kodak. 1966. Kodak Color Dataguide. Basic Books. Rochester, New York: Eastman Kodak. Carson, Rachel L. 1958. The Sea around Us. New Eddington A. S. 1933. The Nature of the Physical York: Mentor Books. World. New York: Macmillan Company. Chakravarti, Tapo Nath. 1978. The Universe of Colour: Edman, Irwin, ed. 1936. The Philosophy of Santayana. Modern Western and Ancient Indian Perspectives. New York: Random House. Calcutta: Putul Chakravarti. Efron, Robert. 1969. "What is Perception?" In Boston Chandler, Albert. 1934. Beauty and Human Nature. Studiesin the Philosophy of Science, edited by New York: n.p. Robert S. Cohen and Marx W. Wartofsky, Vol. 4. Chapman, Frank M. [1895] 1966. Handbook of Birds Holland: D. Reidl. ofEastern North America. New York: Dover T. S. Eliot. 1957. In On Poetry and Poets. New York: Publications. Noonday Press. Chevreul, M. E. [1845] 1980. The Principles of . 1964. Knowledge and Experience in the Phi Harmony and Contrast of Colors. New York: Gar- losophy of F. H. Bradley. New York: Farrar, 334 Bibliography

Straus and Company. Arts as Combinatorial Mathematics, or How . 1967. Poems Written in Early Youth. to Compose Like Mozart with Dice." Scientific New York: Farrar, Straus and Giroux. American 231 (December): 132-36. Emerson, Ralph Waldo. [1836] 1969. In Selected Gauss, Charles Edward. 1966. The Aesthetic Poetry and Prose, edited by Reginald L. Cook. Theories of French Artists. Baltimore: Johns New York: Holt Rinehart Winston Inc. Hopkins Press. Evans, Ralph M. 1948. An Introduction to Color. Geiger, L. [1871] 1880. Contributions to the New York: John Wiley. History and Development of the Human Evans-Wentz, W. Y., ed. [1927] 1966. The Race. Translated by D. Ascher. London. Tibetan Book of the Dead. New York: Oxford Originally published as Zur University Press. Entwicklungsgeschichte der Menschheit. Faulkner, Waldron. 1972. Architecture and Geschwiild, Norman. 1969. "Anatomy and Color. New York: Wiley-Interscience. Higher Functions of the Brain." In Boston Fechner, Gustav Theodor. [1860] 1966. Elements Studies in the Philosophy of Science, edited of Psychophysics. New York: Holt, Rinehart by Robert S. Cohen 4nd Marx W. Wartofsky, and Winston. 4:98-136. Holland: D. Reidl. Feigl, Herbert. 1956. Minnesota Studies in the Gerstner, Kate. 1986. The Forms of Color. Philosophy of Science, vol. 1, edited by Cambridge, Massachusetts: MIT Press. Herbert Feigl and Michael Scriven. Gibson, James J. 1950. The Perception of the . 1958. "The 'Mental' and the 'Physical."' Visual World. New York: Houghton-Mifflin. In Concepts, Theories and the Mind-Body . 1968. "What Gives Rise to the Perception Problem, vol. 2 of Minnesota Studies in the of Motion?" Psychological Review 75 (July): Philosophy of Science, edited by Herbert Feigl, 335-46. Michael Scriven and Grover Maxwell, 385. Gjertsen, Derek. 1986. The Newton Handbook. Minneapolis: University of Minnesota Press. London and New York: Routledge and Kegan Feigl, Herbert, Michael Scriven, and Grover Paul. Maxwell, eds. 1958. Concepts, Theories and Gladstone, W. E. 1858. Studies on Homer and the Mind-Body Problem. Vol. 2 of Minnesota the Homeric Age. 3 vols. Oxford: Oxford Studies in the Philosophy of Science. University Press. Minneapolis: University of Minnesota Press. . 1877. The Colour Sense. Field, George. 1858. Rudiments of the Painter's Gleason, H. A., Jr. 1961. An Introduction to Art; or, A Grammar of Colouring. London: Descriptive Linguistics. New York: Holt, John Weale. Rinehart and Winston. . 1885. Field's Chromatography, A Gogh, Vincent van. 1958. The Complete Letters. Treatise on Colours and Pigments for the Use Greenwich, Conn.: New York Graphic of Artists. Modernized by J. Scott Taylor. Society. London: Winsor and Newton. Gombrich, Ernst. 1956. Art and Illusion. Fielding, Raymond. 1968. The Technique of Princeton: Press. Special Effects Cinematography. New York: Goodman, Nelson. 1957. The Structure of Hastings House. Appearance. New York: Bobbs-Merrill. Focillon, Henri. 1948. The Life of Forms In Art. . 1968. Language of Art. New York: New York: Wittenborn, Schultz. BobbsMerrill. Frank, Philip G., ed. 1956. The Validation of Govindjee and Rajni Govindjee. 1974. "The Scientific Theories. Boston: Beacon Press. Absorption of Light in Photosynthesis." Freud, Sigmund. [1905] 1960. Jokes and Their Scientilic American 231 (December): 68-82. Relation to the Unconscious. New York: W. Greenough, Horatio. 1962. Form and Function: W. Norton Company. Remarks on Art, Design and Architecture. Fry, Roger. 1958. Cezanne. New York: Noonday Berkeley: University of California Press. Press. Gregory, R. L. 1966. Eye and Brain: The Fuchs, Ernst. [1908] 1924. Textbook Psychology of Seeing. New York: McGraw-Hill ofOphthalmology. Translated by Alexander Book Company. Duane, M.D. Philadelphia: J. B. Lippincott Gribbin, John. 1984. In Search of Schrodinger's Company. Cat: Quantum Physics and Reality. New York: Gallant, Roy A. 1961. The ABC's of Astronomy: Bantam Books. An Illustrated Dictionary. New York: Grohmann, Will. 1958. Kandinsky: Life and Doubleday and Company. Work. New York: Abrams. Grosser, Maurice. 1959. The Painter's Eye. New Gardner,Martin.1974."Mathematical Games:The York: Bantam Books. Guilman, Maxine Kraut.1967."Writings on Color Bibliography 335

in Sixteenth Century Italian Art Theory." Wheaton, Ill. Theosophical Publishing House. Master's thesis, Department of Art History, Hofmann, Hans. 1948. Search for the Real. Columbia University. Andover, Mass.: Addison Gallery. Guptill, Arthur L. n.d. Color Manual for Artists. Howard, Frank. 1838. Colour as a Means of Art, New York: Van Nostrand Reinhold Company. Being an Adaptation of the Experience of Hahn, Hans. [1933] 1966. "Logic, Mathematics Professors to the Practice of Amateurs. and Knowledge of Nature." In 20thCentury London: J. Thomas. Philosophy: The Analytic Tradition, edited by Hoyle, Fred. 1957. Astronomy: A History of Morris Weitz. New York: The Free Press. Man's Investigation of the Universe. New Hamlin, D. H. 1949. "The Ridgway Color York: Crescent Books. Standards with a Munsell Notation Key." Imperial Chemical Industries. 1971. ICI Journal of the Optical Society of America ColourAtlas. London and Cleveland: Imperial 39:592. Chemical Industries. Harris, Moses. [1766] 1963. The Natural System International Printing Corporation. 1935. Three of Colors by Moses Harris. Edited by Faber Monographs on Color. New York: Research Birren. New York: Whitney Library of Design. Laboratories of the International Printing Harrison, Bernard. 1973. Form and Content. Corporation. Oxford: Blackwell. Ittelson, W. H. 1952. The Ames Demonstrations Hatt, Joseph Arthur Henry. 1925. The Colorist: in Perception. Princeton: Princeton University Designed to Correct the Commonly Held Press. Theory That Red, Yellow, and Blue Are the Itten, Johannes. 1963. Design and Form: The Primary Colors, and to Supply the Much Basic Course at the Bauhaus. New York: Needed Easy Method of Determining Color Reinhold. Harmony. Together with a System of Color Ivins, William M., Jr. 1946. Art and Geometry: A Nomenclature and Other Practical Study in Space Intuitions. Cambridge: Information for Artists and Workers or Harvard University Press. Designers in Color. New York: D. Van Jacobson, Egbert. 1942. The Color Harmony Nostrand Company. Manual and How to Use It. Chicago: Color Hattersley, Ralph. 1970. "Do You Have Eyes in Laboratories Division, Container Corporation Your Skin?" Popular Photography, March, 55- of America. 58. . 1948. Basic Color: An Interpretation of Hawking, Stephen W. 1988. A BriefHistory of the Ostwald Color System. Chicago: Paul Time: From the Big Bang to Black Holes. New Theobald. York: Bantam Books. Jenkins, Francis A., and Harvey E. White. 1957. Hay, D. R. 1845. The Principles ofBeautyin Fun damentals of Optics. New York: McGraw Colouring Systematized. London: William Hill Book Company. Blackwood and Sons. Judd, Deane B. 1979. Contributions to Color Helmholtz, Hermann von. [1909] 1962. Science. Edited by David L. MacAdam. Special Helmholtzs Treatise on Physiological Optics. publication 545 prepared for the National Ed. James P. C. Southall. 2 vols. New York: Bureau of Standards. Dover Books. Judson, J. A. V. 1935. A Handbook of Colour: A Henri, Robert. 1923. The Art Spirit. Compiled by Textbook for Students, Teachers of Art, and Margery Ryerson. Philadelphia: J. B. All Interested in Color. Leicester: The Dryad Lippincott. Press. Hering, Ewald. [1905-20] 1964. Outlines of a Katz, David. 1935. The World of Color. London: Theory of the Light Sense. Translated by Leo K. Paul, Trench, Trubner and Company, Ltd. M. Hurvich and Dorethca Jameson. Keersmaecker, A. de 1883. Les sees des Cambridge: Harvard University Press. couleurs chez Homere. Heron, Patrick. 1955. "Space in Color." Arts Kelly, Kenneth L. 1965. "A Universal Color Digest 29 (March 15): 8-11. Language." Color Engineering 3 (MarchApril). Herrington, Donald E. 1968. How to Read Reprinted by National Bureau of Standards, Schematic Diagrams. Indianapolis: Howard NBS Circular 553. W. Sams and Company, Inc. Koffka, K. 1922. "Perception: An Introduction to Hickethier,Alfred.1963.Color Mixing by Numbers Gestalt-tbeorie." Psychological Bulletin .New York:Van Nostrand Reinhold Company. 19:551-85. Reprinted, 1966, in Experiments Hiler,Hilaire.1942.Color Harmony and Pigments in Visual Perception, edited by M. D. Vernon. Chicago: Favor, Ruhl and Company. Baltimore: Penguin Books. Hodson,Geoffrey.1976.Music Forms:Superphysi- . 1935. Principles of Gestalt Psychology. cal Effects of Music Clairvoyantly Observed. London: Kegan, Paul. 336 Bibliography

Kohler, Ivo. 1972. "Experiments with Goggles." Maxwell, James Clerk. [1890] 1965. Scientific In Altered States of Awareness, edited by Papers. Edited by W. D. Niven. 2 vol. New Scientific American. San Francisco: W. H. York: Dover Press. Freeman and Company, 1972. Moholy-Nagy, L. [1921] 1947. Vision in Motion. Kohler, Wolfgang. 1947. Gestalt Psychology. Chicago: Paul Theobald. New York: Mentor Books. Mueller, Conrad G., and Mae Rudolph. 1966. Koyre, Alexander. 1965. Newtonian Studies. Light and Vision. New York: Life Science London: Chapman and Hall. Library. Ladd-Franklin, Christine. 1929. Colour and Munsell, A. H. [1905] 1961. A Color Notation: An Colour Theories. New York: Harcourt, Brace Illustrated System Defining All Colors and Their and Company. Relations by Measured Scales of Hue, Value and Land, Edwin. [1959] 1961. "Color Vision and the Chroma. Baltimore: Munsell Color Company. Natural Image." In Symposium on Visual . 1942. Munsell Book of Color. Baltimore: Problems of Color. New York: Chemical Munsell Color Company. Publishing Company. . 1969. A Grammar of Color. Edited by Leger, Fernand. 1946. "Modern Architecture and Faber Birren. New York: Van Nostrand Color." Translated by George L. K. Morris. Reinhold Company. American Abstract Artists exhibition catalog. Nast, Seyyed Hossein. 1978. An Introduction to New York: Ram Press. Islamic Cosmological Doctrines. Boulder, Leland, Nita. 1985. Exploring Color. Cincinnati: Colo.: Shambhala. North Light Publishers. National Bureau of Standards. Inter-Society Leonardo da Vinci. [1802] 1957. The Art of Color Council. n. d. a. The ISCC-NBS Method Painting. Translated by John Francis Rigaud. of Designating Colors and a Dictionary of New York: Philosophical Library. Color Names. NBS circular 553. . 1939. The Notebooks of Leonardo da . n.d. b. The ISCC-NBS Color-Name Vinci. Translated by Edward MacCurdy. New Charts illustrated with Centroid Colors. York: Reynal & Hitchcock. Supplement to NBS circular 553. Lewitt, Sol. 1975. Lines and Color. Zurich: Newton, Sir Issae. [1704] 1952. Opticks, or a Annemarie Verna. Treatise of the Reflections, inflections & Linton, W. "A List of the Principal Colours Used Colours of Light. 4th ed. New York: Dover in Painting." Art Journal (London) 2:271. Publications. Loeb, Leonard B., and Arthur S. Adams. 1933. Optical Society of America. 1953. The Science of The Development of Physical Thought. New Color. New York: Thomas Y. Crowell York: John Wiley. Company. Luckiesh, Matthew. 1938. Color and Colors. Osborne, Roy. 1980. Lights and Pigments: New York: Van Nostrand. Colour Principles for Artists. New York: Maerz, A., and M. Rea Paul. 1930. A Dictionary Harper and Row. of Color. New York: McGraw-Hill Book Ostwald, Wilhelm. 1906. Letters to a Painter. Company. Translated by A. W. Morse. New York: Gin n Magnus, H. 1877. Die geschichtliche Entwicklung & Company. ties Farbensinnes. Leipzig. . 1916. Die Farbfibel. Leipzig: Unesma. . 1880. Untersuchungen uber den [1916] 1969. The Color Primer. Edited by Farbensinn tier Naturvolker. Jena. Faber Birren. New York: Van Nostrand Magnus, Rudolph. 1949. Goethe as a Scientist. Reinhold Company. New York: Henry Schuman. . 1931-33. Color Science. Translated by J. Malevich, Kasimir. 1971. "An Attempt to Scott Taylor. 2 vols. London: Winsor and Determine the Relation between Colour and Newton. Form in Painting." In Malevich, Essays on Art 1932-35. The Ostwald Colour Album. 1915-1933. New York: Wittenborn. Arranged by J. Scott Taylor. London: Winsor Martindill, M. G. 1988. "A European Eye on and Newton. Luminescence: Fluorescent and Ouspensky,P.D.1947.Tertium Organum:The Phosphorescent Colours for the Visual Third Canon of Thought,A Key to the Enigmas of Artist." Leonardo: Journal of the International the World. New York:Alfred A. Knopf. Society for the Arts Sciences and Technology Perkins, Moreland. 1971. "Sentience." The 21 (2): 187-90. Journal of Philosophy 67 ( June 3 ): 329- 37. Matthaei, Rupprecht, ed. 1971. Goethe's Color Pitcher, George, ed. 1969. Wittgenstein, The Theory. Translated by C. L. Eastlake, with Philosophical Investigations: A Collection of notes by R. Matthaei. New York: Van Critical Essays. New York: Doubleday, Nostrand Reinhold Company. Anchor Books. Pitt, F.H.G.1935."Characteristics of Dichromatic Bibliography 337

Vision." In Report of the Committee on the Natural Knowledge. Physiology of Vision, xiv. Special report series, Rucker, Rudy. 1984. The Fourth Dimension: A Guided no. 200. London: Medical Research Council. Tour of the Higher Universes. Boston: Houghton Planck, Max. n.d. A Survey of Physics: A Collection Mifflin Company. of Lectures and Essays. Translated by R. Jones Russell, Bertrand. 1948. Human Knowledge, Its Scope and D. H. Williams. New York: Dutton. and Limits. New York: Simon and Schuster. Plateau, J. 1873. Statique experimentale et theorique Sapir, Edward. 1921. Language: An Introduction to the des Liquides soumis aux seines Forces Study of Speech. New York: Harcourt, Brace and moMculaires. Paris: Gauthier-Villars. World. Plochere, Gladys and Gustav. 1948. Plochere Color . 1949. Selected Writings in Language, Culture, System: A Guide to Color and Color Harmony. and Personality. Berkeley: University of California Los Angeles: Fox Printing Company. Press. Poling, Clark V. 1986. Kandinsky's Teaching at the Sargent, Walter. [1923] 1964. The Enjoyment and Use Bauhaus: Color Theory and Analytical Drawing. of Color. New York: Dover Publications. New York: Rizzoli. Sartre, jean-Paul. 1950. The Psychology of Imagination. Portmann, Adolph, et al. 1977. Color Symbolism: New York: Philosophical Library. Six Excerpts from the Eranos Yearbook, 1972. Schillinger, Joseph. 1941. The Schillinger System of Dallas: Spring Publications. Musical Composition. New York: Carl Fischer, Inc. Ray, V. F. 1953. "Human Color Perception and 1948. The Mathematical Basis of the Arts. Behavioral Response." Transactions of the New New York: Philosophical Library. York Academy of Sciences 16:98-104. Scripture, E. W. 1895. Thinking, Feeling, Doing. New Rewald, John, 1943. . New York: York: Flood and Vincent. Wittenborn. Segall, Marshall H., Donald T. Campbell, and Melville J. Rhees, R. 1969. "Can There be a Private Language?" Herskovits. 1966. The Influence of Culture on In Wittgenstein, The Philosophical Investigations: Visual Perception: An Advanced Study in A Collection of Critical Essays, edited by George Psychology and Anthropology. New York: Pitcher. New York: Doubleday Anchor Books. Bobbs-Merrill. Ridgway, R. 1886. A Nomenclature of Colors for Skinner, B. F. 1972. Beyond Freedom and Dignity. New Naturalists. Boston: Little, Brown. York: Bantam Books. . 1912. Color Standards and Color Smith, Charles N. 1965. Student Handbook of Color. Nomenclature. Baltimore: Hoen. New York: Van Nostrand Reinhold Company. Rivers, W. H. R. 1901. "Primitive Color Vision." Societe Francaise des Chrysanthemistes and R. Popular Science Monthly 59:44- 58. Oberthur. 1905. Repertoire de Couleurs pour Aider Robinson, Daniel Sommer. 1947. The Principles of a la Determination des Fleurs, des Feuillages et des Reasoning: An Introduction to Logic and Fruits. Paris: Libraire Horticole. Scientific Method. New York: Appleton-Century Soury, J. 1878. De 1'evolution historique du sens des Crofts, Inc. couleurs. Rood, Ogden N. [1879] 1913. Student's Textbook of Standard Dictionary of Folklore, Mythology, and Color, or Modern Chromatics with Applications to Legend. 2 vols. New York: Funk and Wagnalls. Art and Industry. New York: D. Appleton & Stokes, Adrian. 1937. Colour and Form. London: Faber Company. and Faber Ltd. Rood, Roland. 1941. Color and Light in Painting. Strathmore Paper Company. 1921. A Grammar of Edited by George L. Stout. New York: Columbia Color. Westfield, Massachusetts: Strathmore Paper University Press. Company. Rorschach, Hermann. 1964. Psychodiagnostics. Sullivan, J. W. N. 1965. The Limitations of Science. Translated by Paul Lemkau, M.D. 6th ed. New New York: Van Nostrand Reinhold Company. York: Grune & Stratton. "Surangama Sutra, The". 1942. In The Wisdom of China Ross, Denman Waldo. 1919. The Painter's Palette: A and India, edited by Lin Yutang. New York: Random Theory of Tone Relations, An Instrument of House. Expression. New York: Houghton Mifflin. Swedenborg, Emanuel. [1758] 1965. Heaven and Its Royal Society of London for Improving Natural Wonders and Hell. Translated by John C. Ager. New Knowledge. Philosophical Transactions: giving York: The Citadel Press. some Accompt of the present undertakings, Sweeney, James Johnson. 1945. Stuart Davis. New studies and labours of the Ingenious in many York: Museum of Modern Art. considerable parts of the world. 1664-1832. Lon- Symposium on Visual Problems of Color. 1961. New don:The Royal Society of London for Improving 338 Bibliography

York: Chemical Publishing Company. Walls, Gordon Lynn. 1942. The Vertebrate Eye. Taylor, H. D., L. Knoche, and W. C. Granville. 1950. Bloomfield Hills, Mich.: Cranbrook Press. Descriptive Color Names Dictionary. Chicago: Whistler, James McNeil. 1890. The Gentle Art of Making Container Corporation of America. Enemies. New York: John W. Lovell Company. Taylor, Joshua C. 1961. Learning to Look: A Handbook Whitehead, Alfred North. [1925] 1953. Science and the for the Visual Arts. Chicago: University of Chicago Modern World. New York: Mentor Books. Press. Whittaker, Sir Edmund. [1910] 1951. A History of the Taylor, J. Scott. 1935. A Simple Explanation of the Theories of Aether and Electricity. 2 vols. New York: Ostwald Colour System. London: Winsor and Harper Torchbooks. Newton Ltd. Whorf, Benjamin Lee. [1956] 1967. Language, Thought Textile Color Card Association of the United States, Inc. and Reality: Selected Writings. Edited by John B. 1941. Standard Color Card ofAmerica. New York: Carroll. Cambridge: MIT Press. Textile Color Card Association. Wilson, Anton. 1970. "Film Feedback." Filmmaker's Titchener, E. B. 1916. "On Ethnological Tests of Newsletter (September): 32. Sensation and Perception with Special Reference to Wittgenstein, Ludwig. 1958. The Blue and Brown Tests of Color Vision and Tactile Discrimination Books: Preliminary Studies for the "Philosophical Described in the Reports of the Cambridge Investigations." New York: Harper Torchbooks. Anthropological Expedition to the Torres Straits." 1967. Remarks on the Foundations of Proceedings of the American Philosophical Society Mathematics. Edited by G. H. von Wright, R. Rhees, 55:204-36. and G. E. M. Anscombe. Cambridge: MIT Press. Tonks, O. J. 1934. Colour Practice in Schools: A Graded Wright, W. D. [1944] 1969. The Measurement of Colour. Course in Colour Seeing and Using for Children New York: Van Nostrand Reinhold Company. Between the Ages of Five and Fifteen. London: Winsor . 1962. Proceedings of the Conference on and Newton Ltd. Optical Instruments and Techniques, 1961. London: Triandis, H. C. 1964. "Cultural Influences on Cognitive Chapman and Hall. Processes." In Advances in Experimental Social . 1968. The Rays Are Not Coloured: Essays on Psychology, vol. 1, edited by L. Berkowitz. New the Science of Vision and Colour. New York: York: Academic Press. American Elsevier Publishing Company. Vanderpoel, Emily Noyes. 1903. Color Problems: A Wurtman, Richard J. 1975. "The Effects of Light on the Practical Manual for the Lay Student of Color. New Human Body." Scientific American 233 York: Longmans, Green and Company. (July): 68-77. van Gogh, Vincent. See Gogh, Vincent van. Young, Thomas. 1802. "On the Theory of Light and Vasari, Giorgio. [1550] 1957. Lives of the Artists. Colours." Philosophical Transactions of the Royal Translated by E. L. Seeley. New York: Noonday Society 92:20-71. Press. Zeishold, Herman. 1944. "Philosophy of the Ostwald Vernon, M. D., ed. 1966. Experiments in Visual Color System." journal of the Optical Society of Perception. Baltimore: Penguin Books. America 34 (July). Wald George. 1961. "Retinal Chemistry and the Zimmer, Heinrich. 1946. Myths and Symbols in Indian Physiology of Vision." In Symposium on Visual Art and Civilization. Edited by Joseph Campbell. Problems of Color. New York: Chemical Publishing New York: Pantheon Books. Company. Wallace, Florence Elizabeth. 1927. Color in Homer and Ancient Art. Smith College Classical Series, no. 9. Northhampton, Mass.: Smith College. Index

Achromaticity, 54 symbolic, 120, 122 Maxwell disc test, 304-306 Additive/subtractive theory, Tyndall, 62 visual Munsell wheel, 310-311 323-328 affinity for, 54 as negation, 303-304 Agnosias. See Visual agnosias Bosons, 248-249 Ostwald wheel, 310-311 Airglow, 49 Boyle, Robert, 62 cause of product color test, 306-308 Allen, Grant, 10 color theory, 74 wavelengths in light, Anomaloscope, 199 Brain damage, relation to color 308-310 Aristotle, 3, 62-63 light effects information, 40 as a continuum, 11-12 versus color effects, 103 Art, Brown, cause of, 85-86 created by light, 62 color and form in, 224-229 describing, 257-258 Artists, use of light, 114-120 differentiating, 234 Astronomy, relation to color, doubting, 34-36 58-61 Cantor, Georg, 11 experience, 41 Carnap, Rudolph, 13, 14 fields theorem weaknesses, 23 constancy, 181-182 Bassa language, 201-203 Chapman's color swatches, displacement, 180-181 Bible, 34 24-25 noninterpenetrability, Birren, on color clash, 279-280 Charcot, jean-Martin, 197-198 179-180 Black, 88-91 Color. See also specific colors perimeters, 183-184 as an absolute, 53-54 achromatic, 54, 88 separating form and, Goethe's views, 208 as light tristimulus values, 95. See 177-179 absorber, 91 relation to also Black; White in art, surface, 182-183 nothingness, 50-52 symbolic, 224-229 cause of, 73-78 surface geometry, 183 120-122 causing light, 97-98 union with form, 175-177 Blue chimpanzee studies, 41 form and, 237-239 ambiguous names for, 265 chroma, 14 clash, 276-280 naming, 239-244 astronomy related to, 59 coding, by wavelengths, games, by Ludwig Wittgen- describing ultraviolet rays, 85 93-94 complementary, stein, 29-31 Goethe's views, 208 301-303 harmony and, generic, 14-15 moderate, ISCC chart, 19 311-312 hell associated with, 107 hue, 14, 187192

339 340 Index

Ibo classifications, 205 oil paints, 20 value, 14 versus Kandinsky's scale, 207 origin of, 4-5 noncolor, 38 wheels, identifying, visual agnosias, ostensive definition, 4 80-81, 266-267 37-41 pathologies named after, 8 Color blindness. See illusion related t( , 135-136 people named after, 8 Tristimulus theory industrial standards, 20 pigments, 20 Colorlessness, 92 intermediary ranges, 11 poetic applications to, Color standards, 255-256 language, 13, 26 23-24 Copenhagen interpretation of Bassa, 201-203 rules for, 7 quantum theory, 67 classes, 27-28, 60 scientific applications to, Corpuscular theory, 67 comparisons, 32-33 23-24 Cube illusion, 148-149 concepts, 28-29 similes, 9, 24-25 Ibo, 205-209 visual comparisons, 17-18 Daltonism, 128 mistakes, 35 in nature, 209-212 Darkness, symbolic, 108-109 Shona, 203-204 nonvisual experiences with, da Vinci, Leonardo, on light in sorting criteria, 29-31 126-131 relation to color, 48, 60 vagueness of, 27 notational systems, 251-253 Descartes, 63 impossible locating in light, 231-232 ostensive definition and, figures, 150-151 logic and, 33-34 256 Dichromatism, 216 lower organisms relation to, numbers and, 259-261 DIN 6164, 24 41-42 origin of word, 106-107 Dye technology, 9-10 measuring instruments, 128 as a percept, 31-33, 80-81 aniline dyes, 10 memory, 42-43 perception, 27-28 mixture, 313-316 in perspective, 57-61 Eddington, A. S., 102 color wheel to predict, photography, 219 Electromagnetic spectrum, 316-319 primary, 293-294 93-94 lights, 319-320 additive/subtractive theory, Electromagnetic theory, 82-83 mixed colors without, 316 326 Electromagnetic wave scale, optical, 320 mixing, 296 80, 84 pigment mixing, 319 Young's theory, 294-295 Elements, chemical properties, teaching, 321-322 purity, 297 G9-70 muddiness, 297 secondary, 294, 296 Eliot, T. S., 23 light symbolism names, 3, 43-44 sensibility, 10 used by, 110 "The Death of ambiguous, 262-267 as something and nothing, Saint Narcissus," 35-36 chemical names as, 55-57 Euclidean geometry, 152-156 258-259 spectral, 90 Eye, seeing color, 232-233 classes of, 21 wavelengths, 324 continuousness of color staining, 10 Feigl, on nonvisual seeing, and, 12-16 standardizing, 17 127-128 derived from objects, 5-7 standards, 61-62 Fermions, 248-249 dictionaries of, 18 subjectivity, 234-236 Fog, 50 forgetting, 41-42 swatches, 18, 257 Fuchs, Ernst, 199 ish endings, 22 symbolic, 30, 120-125 Fuchsia, ambiguous names for, metaphors, 24-25 future of, 124-125 264 nonsensical, 72 systems, 272-276 objects named after, 8-9, Munsell versus Ostwald, Ganzfeld, 50 25 276 Gathercoal, color classes, 20-22 ohmic value of carbon tertiary, 294, 297 Geometry, of visual space, resistors and, 8 undifferentiated, 49-50 151-152 Index 341

Geschwind, on visual agnosia ISCC. See International Society Magenta, ambiguous names research, 39-41 Color Council for, 264 Gladstone, William Ewart, on Ivins, William M., 33 Maxwell, James Clerk, disc test, 304-306 color language, 193-200 Maxwell triangle, 214 Goethe, views on blue and black, 208 Memory, relation to color, 42-43 Gray metameric, 130 scale, Judd's color modifiers, 22-23, logarithmic, 281 Metamerism, 129-130, 219-222 217-219 Michelangelo, lightness and experimental procedures, ISCC-NBS color names and, 282-284 geometric darkness symbolism used 24 by, 112 progression, 286 norms, 286 practical, 287 Monochromatism, 214-215 Moses, use of halos, 119-120 Language. See Color, language Muller-Lyer arrows, 146-147 Last Supper, use of halos, Grayness, 49 Munsell, Albert H., 6 118-119 Greek logic, 33-34 Munsell color chart, 12, 14, Law of contradiction, 32-33 Green, symbolic, 123 253-255,266-267 Light. See also Color, causing Gregory, R. L., and optical complementary colors, light absorbers, 91 illusions, 146 302-303,310-311 ISCC appearance, 100-104 classification compared to, biblical references to, 24 109-110 Hahn, Hans, 29 cause of, 73-78 Hallucinations, 144-145 National Bureau of Standards, dependent on color, 103 20 color classes, 21-22. See Halos, 118-120 infrared, 83 Harris, Moses, 7 also International Society instantaneous propagation, Color Council Heat scale, 83-84 99-100 Henley, William Ernest, Newton, Isaac movement of, 104, 244-245 darkness symbolism used color wheel, 80-81 multiple reflectance, by, 111 criticism of Hooke, 64-65, 68 104-105 Henna, 19-20 nature of light, 66-70 opacity, origin of word, 106 Hooke, Robert, 63 70-72 perspective and, 113 Netwon's criticism of, 64-65, translucency, 70-72 artists' use of, 114-120 68 transparency, 70-72 reflectors, 91 Night blindness, 215-216 Hue. See Color, hue; Optical relation to color daylight, mixture Not-colors, 31-33 47-48 deprivation of, 48

excessive, 48 Oil paints, color names used lbo language, 205-209 Hooke's theories, 64-65 hue classifications, 205 for, 20 Newton's theories, 64-66 Optical illusions, 145-147. See Illumination, 35. See also Light night darkness, 48-49 also specific types opacity, 70-72 Indigo, ambiguous names for, Optical mixture, 86-87 translucency, 70-72 265 for brown, 85-86 transparency, 70-72 International Society Color circular diagrams, 81 shadow and, 116 Council circularrelationships,82-84 speed, 99 color classes, 21-22, 252 hue relationships, 79-80 symbolic, 107-110, 114-120 lengthiness of, 24 linear diagrams, 81 time and, 245-248 color modifiers, 22 linear relationships, 82-84 transparency, causes of, standard colors, 19 second continuum, 81-82 91-93 ultraviolet, 84-85 342 Index

ultraviolet rays, 84-85 astronomy related to, 59 color as surface, 171-172 Optical Society of America, 24 symbolic, 120-121, 122 coordinates of vision, Orange ambiguous names for, Reddish orange varieties, 19 172-174 265 symbolic, 123 Rembrandt, lightness and dark form and color, 168-169 Ostwald, Wilhelm, 6 color wheel, ness symbolism used by, form as discontinuity, 266-267 complementary, 116-117 170-171 302-303, 310-311 gray scale, Renoir, use of warm colors, 202 Visual field extension 285-292 Rivers, W. H. R., 5 color, 158-159 Roemer, Ole, 99 dependent on becoming, Rorschach inkblot test, 43-44 159-161 Russell, Bertrand, 4, 31-32 edge of, 162-164 Parmenides, and delusion, form, 158-159 143-144 fragmentation, 164-165 Particle hypothesis, 66-67 Sellers, Wilfred, 35 color in, 166 Perceptionism, 279-280 Shona language, 203-204 holes, 157-158 Perkin, William Henry, 9-10 Solipsism, 137-138 size and, 161-162 Perspective, 52-53, 57-61 Sunlight, 60-61 spaces, 157-158 Pitch scale, 83-84 Supersaturated substance, 86 Visual thinking, 57 Pitt, F. H. G., confusion colors, 217-218 Planck, Maxwell, 68 cause of light Time, and light, 245-248 Wallace, Alfred Russel, on color theory, 75-76 colorless versus Tristimulus theory, 213-214 subjectivity, 235 black distinction, 91-92 Color blindness model, 222 Wave hypothesis, 66-67, 68 Plants, relation to color, 42 dichromatism, 216 Weight scale, 83-84 Plato, 63 monochromatism, 214-215 White as color of light, Pompeiian red, name origin, 27 seeing colors differently, 65 as light absorber, Purple ambiguous names for, 216-219 91 misuse of word, 92 265 symbolic, 122 shades of, 94-96 symbolic, 120-122 Ultraviolet light. See Light, Whitman, Walt, 24 ultraviolet Wittgenstein, Ludwig, 29-31 Rainbows, 62-63 Rays, 66 Reality defining, 137-138 Vermeer, lightness and darkness X rays, 76 illusion related to, 135-136 real symbolism used by, 117 chairs, 138-142 Vermilion, name origin, 27 Red ambiguous names for, Violet, ambiguous names for, Yellow, symbolic, 123-124 263-264 265 Young, Thomas, primary color Visions, 144-145 theory, 294-295 Visual agnosias, 37-41 Visual field complementarity, Zeno, and delusion, 144 167-170 Other Books from Design Press

PRINCIPLES OF VISUAL PERCEPTION-2ND EDITION by Carolyn M. Bloomer

How and why do people's perceptions of a work of art differ? Combining psychology and art theory, artist and educator Carolyn M. Bloomer explores how we construct meaning from visual information and how the physiological aspects of visual perception are influenced by culture and experience. Making reference to the problems of creating and teaching art, Bloomer offers practical applications for the ideas of each chapter with studio projects designed for both classroom and individual practice.

Hardcover $27.45 Book No. 50004

SYMBOLS SIGNS LETTERS by Martin Andersch

Elegantly designed, exquisitely produced, and winner of a gold medal at the Interna tional Book Art Exhibition (Leipzig, 1989), this collection presents the art of calligra phy at its finest and most innovative level. It combines professor of design Martin Andersch's observations on the state of lettering in the twentieth century with extraor dinary work done by students in his seminars. Fine Print magazine says, "It is a beau tiful book, with inspiring pictures and ideas for teachers, students, and laymen . . ."

Hardcover $74.00 Book No. 50006

LOGIC AND DESIGN by Krome Barratt

Perception and communication specialist Krome Barratt examines key principles of design and their relationship to art, mathematics, and science in this unique combination of text and graphics. A sourcebook of ideas from a variety of disciplines, all concerned with innovation and problem solving, it covers such topics as number, ratio and scale, rhythm and harmony, and similarity and contrast, and suggests ways that the designer can creatively apply them.

Paperback $16.60 Book No. 50012 *Prices subject to change without notice.

Ask for Design Press books at your local bookstore or write

Design Press P.O. Box 40 Blue Ridge Summit, PA 17294-0850

TO ORDER TOLL-FREE: 1-800-822-8138