Introduction

Modernity is colorful. And literature of the early twentieth century, coincident with the modern, or modernist, period, is brighter, more colorful, and more visual than literature of the early nineteenth century: 91.5% more colorful. By this time, the fiction and poetry published in Britain, and by extension in much of the Anglophone world, contains more color words, more colorful objects, and more descriptions of color phenomena than it ever did, and measurably so. Color and visual phenomena become not only modes of description, ancillary to plot and narrative arc, but themselves become themes, subject matter, and organizing principles. By computationally modeling color in literature, in this chapter, and by contextualizing these results within the artistic and material history of the period, I will not only show the explosion of color with numeric precision, but I will show the workings of color in literature: it generic affiliations, its historical fluxuations, and its linguistic fingerprint. This chapter thus operates at the nexus of quantitative text analysis, color studies, and phenomenology.

Color has always been one of the most difficult subjects to study. It’s a classic epistemological problem. On the one hand, colors appear to be a stable set of cognitive categories that we learn in primary school: basic properties of the things around us. Upon closer examination, however, various complicating phenomena reveal color to be so conceptually elusive that it ultimately forces us to challenge the basis of knowledge itself. These phenomena include optical illusions, perceptual differences between individuals and speakers of other languages, and the blurred boundaries between imagination, cognition, speech, and writing. What literature, and art, do to this soup of phenomena is only to complicate them further. These issues make computational analysis all the more difficult to perform, and to interpret.

Up until now, color has largely been considered supplemental to a literary work—it colors a passage, or provides color to it, as if it were an afterthought. While color studies is strong in philosophy, philology, and physics, as well as the certain cross-disciplinary areas of the natural sciences, it is taken less seriously in literary studies of late. My contribution is to show, qualitatively as well as quantitatively, the ways in which color, as a site of phenomenological uncertainty, forms a locus at which many of the problems of modernism are engaged. Not only are modernist writers struggling with ways to write the visual experience with chromatic precision, but the tension between text and image is, I will argue, a central mode of this period’s writing. Reading modernist works in terms of color will give us a new perspective on the visuality we have largely taken for granted, and by positioning color at the center of this inquiry, we destablize the canon, and reinforcing it in other ways, by finding the most colorful, and colorless, works in British literary history.

This chapter continues to analyze the eye of modernism, and focus on the aspect of vision handled by retinal cones: color, hue, and brightness. Color forms an important part of the modernist project which aims, reprising Conrad, to make you see: to convey the visual experience to the reader, from imagination to imagination, via textual encoding. This chapter reveals modernist usage of color to be a vital entry-point into description, word-painting, subjectivity, and visual epistemology.

The centerpiece of this study is a quantitative analysis of color, in British fiction and poetry of the late nineteenth and early twentieth centuries. Since the phenomenon of color is so widely discussed, across disciplines—in philosophy, psychology, anthropology, physics, history, linguistics, and to some degree in literary studies—this study will of necessity be interdisciplinary. I use research on color perception from experimental psychology, for example, to inform an algorithmic design, which also inherits methods from color studies in other sciences. Although the primary method I employ here is quantitative, and its scope literary-historical, many of my findings are derived from close readings—ones which inform quantitative methods, and which inform further readings, and so on. This is a process known in computational statistics as Box’s Loop (Box). As statistician David Blei explains it, “we build a model, use it to analyze data, assess how it succeeds and fails, revise it, and repeat” (Blei 203).

The study of color, imagery, and other visual effects in literature has a long history, a fact which is revealed in the breadth of several bibliographies. Sigmund Skard’s 1946 The Use of Color in Literature: a Survey of Research lists over a thousand works, in several European languages, which discuss textual color as it appears in literature (Skard). Skard lists works of psychology, philology, etymology, semasiology, religion, ethnology, folklore, classics, and modern literatures in a dozen or so languages. Robert Doak’s 1974 Color and Light Imagery: an Annotated Bibliography lists nearly five hundred works (Doak). Doak’s categories are similar, but he adds symbolism, heraldry, and proverbs to his collection. The writings about color more generally are even more numerous. My path will be to chart a narrow path through this area, to interrogate the boundary between word and color.

In modernist studies, there have been a number of notable monographs dealing with textual color. Karen Jacobs’s The Eye’s Mind: Literary Modernism and Visual Culture explores “new kinds of visual relations” in American, British, and French modernist texts (Jacobs). Jacobs historices ocularity in modernism, citing “newly skeptical philosophical discourses of vision in the first half of the twentieth century … which led to what Martin Jay hs called a ‘crisis in ocularcentrism’”; the impact of “visual technologies” like photography and film; and the emergence of anthropology and sociology as academic disciplines. Jack Stewart’s Color, Space, and Creativity deals with the interaction of its eponymous concerns, in Virginia Woolf, D.H. Lawrence, Lawrence Durrell, Joyce Cary, and A.S. Byatt. Stewart argues, as I do here, that “Verbal equivalents of color and space are no mere adjuncts to narrative. They are means by which writers build imagined worlds and focalize characters’ sensibilities” (J. Stewart 15). Among Stewart’s primary interlocutors are visual artists like Van Gogh and Kandinsky, who compile new theories of the interactions of color. David Batchelor’s Chromophobia charts what he sees as an aversion to strong colors, among American writers of this era, leading to monochromatics in works such as Heart of Darkness and Moby Dick. His wide-ranging study includes correspondences with architecture and psychedelics, as in Aldous Huxley’s Doors of Perception (Batchelor and Books). Another monograph with a coinage for a title is Chromographia, by Nicholas Gaskill, who studies “inscriptions of color in writing” (Gaskill 7). Gaskill argues that, “whereas traditional empiricism casts chromatic qualities as deceitful overlays cloaking the real world, the combined efforts of psychologists, philosophers, dyers, and paint manufacturers promoted a view of colors as relational phenomena that cut across the reigning categories of subject and object, inside and outside, nature and culture” (Gaskill 8). (This view, of color complicating dichotomies like subject and object, is also that of May Sinclair, as we shall see below.) My work continues that of these scholars, although focused more on the difficulties of translating between text and imagination.

There is a somewhat shorter tradition of quantifying textual color, as well, which I will build from. The most famous study is probably that of William Gladstone, the classicist and British Prime Minister whose Studies on Homer and the Homeric Age (published in 1858, ten years before he became the British Prime Minister) argues that the ancient Greeks couldn’t see the color blue, and that blue is in fact a modern invention (Gladstone). He arrives at this conclusion from a thorough manual tabulation of color words in Homer.

There are also quantitative color word analyses, in linguistics and philology, by Seija Kerttula, whose English Colour Terms: Etymology, Chronology, and Relative Basicness mines literary and linguistic corpora to determine the chronological development of color terms in English; Sigfried Wyler, whose Color and Language: Color Terms in English finds the frequencies of color words in linguistic corpora, as well as their frequencies of use in metaphor; and Anders Steinvall, who in English Colour Terms in Context takes a cognitive-linguistic approach to studying color terms in the corpus Bank of English (Kerttula; Steinvall, English Colour Terms in Context; Wyler).

In literary studies, Élodie Ripoll’s Penser la couleur en littérature: explorations romanesques des lumières au réalisme is a highly effective diachronic study of several centuries of novels, mostly French, which finds not only that literature passes from a “chromatic poverty” (littérature pauvre en couleur) in the eighteenth century, to a nineteenth century whose color is “at the heart of its poetics” (Ripoll 10). Ryan Heuser and Long Le-Khac, of the Stanford Literary Lab, also study color words in this period, with a view towards changing levels of abstraction and concreteness in British novels. They find a “sharp rise” in color word proportions from 1785 to 1900: from 0.04% to 0.19%, more than a 290% increase (Heuser and Le-Khac 22). Most recently, Ted Underwood’s Distant Horizons: Digital Evidence and Literary Change begins with an analysis of a random sample of 347 volumes of fiction, and arrives at a similar result to mine, in fig. 1 below (Underwood 10).

But these quantitative analyses all use narrower definitions of the textual colors they measure. With the exception of Gladstone, whose manual tabulation allowed him to disambiguate senses, these studies typically measure color words according to their lexemes, undisambiguated by word senses. Think of the ambiguity between rose the color, and rose the past tense of the verb to rise. Orange the fruit, and gold the metal are similar examples (although in imagination, there may be little difference, as I argue later). In Heuser and Le-Khac’s appendix, for instance, they list all 50 color words they use, among which are included orange and gold (Heuser and Le-Khac 55). Similarly, Underwood reveals that his list of color words “is long enough to include colors like ‘cerulean,’ but there will always be chromatic experiences it leaves out and emotions that it mistakes for a color when tallying up occurrences of a word like ‘blue.’” (Underwood 9). As I will demonstrate below, color word ambiguity and polysemy are much more problematic than these confessions suggest.

The algorithms I develop below not only address these problems with the counting of color words, but handle the larger problem of extracting a spectrum of visual information from text. In technical terms, it is a probabilstic model for named entity recognition (NER), trained on a weighted synthesis of color-word mappings of the past century, as well as lexicographical data, and crowdsourced, image-based data from across the Internet. But put succinctly, it is an imagination machine: like a human reader, it is given textual input, and returns visual information: in this iteration, it is color and its distributions. Not only does it see the word avocado and imagines #568203, or leisure and imagines #445d60, but is capable of imagining verbs, nouns, and abstract concepts, as well. This is all described in detail in the experiment section below.

My study contributes to the understanding of textual color, not only by showing that modernist literature is colorful, and that it is 91.5% more colorful (that is, with more precision), but by analyzing color words more holistically, to show the way which visual mechanisms operate in literature. So, I am less concerned with the what of this chromatic explosion in modernism, as I am the why and the how. Regarding the why, there are important contexts and backgrounds to this visual turn: the cultural and historical climate of the early twentieth century; intermedial cross-pollinations between writers, visual artists, and filmmakers; literary and aesthetic theories of the day; literary-generic and market considerations; and material histories, such as the sudden commercial availability of cheap synthetic dyes. I begin the chapter with these contexts.

From there, I introduce some problems with color analysis which inform the algorithmic design: the epistemology of color words and their literary/linguistic functions, which complicate the process of encoding this process into a computer program. Two case studies help to illustrate these problems: a study of Virginia Woolf’s To the Lighthouse, and another of James Joyce’s A Portrait of the Artist as a Young Man.

Next, I describe the what: the experiment itself, the design of the algorithm, the data sources. These are not technical details, I must emphasize, but are crucial decisions, informed by the psychological and philosophical research in color studies, and which defines the computational work to follow. Again, in a crucial sense, the algorithm is the criticism.

Finally, as literary scholars, we know that modernism is colorful. But what is missing is the how, and that is where I conclude this chapter: how, precisely, color appears in text, how it hides in text, and what it achieves in its contexts. I show how genres which were growing during this period, like travel novels, and children’s literature, contribute to this explosion of color, along with genres like love stories, psychological fiction, and detective fiction. I show how trends within these genres—what critics have called literary impressionism, for instance—heighten the senses of the narrators, and increase the likelihood of color descriptors. In short, I show the workings of textual color in literature.

The Problem: Modernity’s Explosion of Color

In Virginia Woolf’s meta-essay, “The Decay of Essay Writing,” she argues that her moment in history “has painted itself more faithfully than any other in a myriad of clever and conscientious though not supremely great works of fiction; it has tried seriously to liven the faded colours of bygone ages” (Woolf, Selected Essays). While she does not use “painted” or “colours” exclusively in their literal sense, she gives the distinct impression that modernity is brighter, and more colorful, than the previous age. The first of my analyses puts this hypothesis to the test, inferring colors from a corpus of hundreds of thousands of novels and poems, in order to answer the question of whether twentieth century writers are more colorful, or more descriptive, than their predecessors. They are.

There is a tremendous increase in the proportions of color words, color expressions, and colorful objects, in British literature at the turn of the twentieth century. Fig. 1 shows the proportions of these colors, in hundreds of thousands of novels and poems, and how they correlate with their authors’ years of birth. Since the generation writing in the high-modernist decades of the 1910s and ’20s is born in the 1870–1900 range, give or take a decade, we see a huge jump in color during this period. This is Woolf’s generation quite literally “livening the faded colours of bygone ages.”

I use the word generation advisedly, as does Woolf. The author’s year of birth might seem like a strange choice of a variable, for an analysis of this sort, since virtually all quantitative studies of color words use publication dates. But as I will explain below, there are strong generational currents that make this choice necessary. Woolf herself divides then-living writers into “Edwardian” and “Georgian” camps, as we shall soon see.

These trends are the results of the complex quantitative analysis described below, which analyzes a very large corpus, using a deep imagination algorithm to identify not only direct expressions of color, but implied visual information, as well. Let’s begin by examining the historical context to this phenomenon: the trends in literary and art history, and social / material conditions for this increase in color.

On or About December, 1910: a King, an Art Exhibition, and a Comet

In her oft-cited essay from 1924, “Mr Bennett and Mrs Brown,” Virginia Woolf famously asserts that December, 1910, was a turning-point for human character:

“My first assertion is … that everyone in this room is a judge of character. Indeed it would be impossible to live for a year without disaster unless one practised character-reading and had some skill in the art. … And now I will hazard a second assertion … that in or about December, 1910, human character changed.” (Woolf, Collected Essays by Virginia Woolf. Vol. 1 320)

The phrase “in/on or about December, 1910” appears over 1,192 times in publications since 1924,Mostly since 1980. I derive this total from raw data from this 5-grams archive, via Google Books.

and is the title of a monograph on the early Bloomsbury group (Stansky). In it, Peter Stansky cites a number of epoch-making events that straddle 1910. Among them are: the death of King Edward VII, signaling the end of the Edwardian era; Roger Fry’s London exhibition, “Manet and the Post-Impressionists”; and the passing, in early 1910, of Halley’s Comet. Each of these is worth addressing, not just as significant events of the new decade, but as events which contributed to the change in the literary ocularities that we see manifested in the figures above.

First, let us consider the end of the Edwardian era. The death of the “rich and vulgar” Edward VII, in May 1910, along with the coronation of George V in 1911, while less culturally significant than other events, nonetheless provided for many Britons a useful shorthand for circumscribing an era of extravagance (Stansky 1).Paul Thompson’s study of the Edwardians finds that “the top 1 per cent of Edwardians … owned 69 percent of the national capital,” a wealth, and an inequality, which “was the highest in modern British history and probably then the highest in the western world” (Thompson 2). We will return to this later, when dealing with material conditions of commercial pigmentation.

Woolf’s essay begins by dividing early twentieth century writers along these lines: “Mr. Wells, Mr. Bennett, and Mr. Galsworthy I will call the Edwardians; Mr. Forster, Mr. Lawrence, Mr. Strachey, Mr. Joyce, and Mr. Eliot I will call the Georgians” (Woolf, Collected Essays by Virginia Woolf. Vol. 1 320). Woolf’s list of British writers is very nearly in chronological order, by date of birth. The most senior of the “Edwardians,” H. G. Wells, was born in 1866, quickly followed by Arnold Bennett and John Galsworthy in 1867, while the “Georgian” E. M. Forster was born a decade later in 1879, followed by Lytton Strachey in 1880, James Joyce in 1882 (as with Woolf herself), D. H. Lawrence in 1885, and T. S. Eliot in 1888.

This generational difference—the motivating factor behind the analysis shown in fig. 1—Woolf sees as manifest in an internal cohesion of the writers’ works. While among the Georgians, or, in a proto-modernist work like Tristram Shandy, Woolf argues, “everything was inside the book, nothing outside,” the Edwardians, in contrast, “were never interested in character in itself, or in the book in itself. They were interested in something outside. Their books, then, were incomplete as books, and required that the reader should finish them, active and practically, for himself” (- Woolf, Collected Essays by Virginia Woolf. Vol. 1 327).

Woolf’s concern for “character in itself” and “the book in itself” hearkens back to the autoteleological aphorism of Aestheticism, “l’art pour l’art,” or “art for art’s sake.” (Aestheticism, the artistic and literary movement of the 1890s, was born of the late Victorians, a generation for whom modernists / “Georgians” felt affinity.) In fact, we might well call Woolf’s essay “Mrs. Brown for Mrs. Brown’s Own Sake.” But what is more important is the mode in which this autotelos is enacted: sight. Sight is useful, or useless, unto itself: it is the progenitor of action, and precedes it.

In comparing Wells, Bennett, and Galsworthy, by imagining the way each novelist would treat Mrs. Brown, Woolf’s train car neighbor, Woolf uses visual and ocular metaphors to contrast their writing styles: Wells would “project a vision on the window-pane” of a utopian world without Mrs. Brown. “And what would Mr. Galsworthy see?”, Woolf asks (- Woolf, Collected Essays by Virginia Woolf. Vol. 1 327). (Crucially, this is not “what would Mr. Galsworthy write,” but what would he see.) He would see a symptom of a failing society, she replies. Bennett, however, “would keep his eyes in the carriage” (- Woolf, Collected Essays by Virginia Woolf. Vol. 1 328). Later, in describing Strachey’s “against the grain” biography Queen Victoria, Woolf argues that “Mr. Strachey has had to open our eyes before he has made us see” (- Woolf, Collected Essays by Virginia Woolf. Vol. 1 335).

This “make us see” is the very same which Conrad identifies as his goal, in the preface of The Nigger of the Narcissus, which I quote in the introduction. It is the goal of the novelist, according to Conrad, and to Woolf. In fact, Woolf acknowledges Conrad, in this essay, as one of the only exemplary writers available to the 1910 generation, for his vision—that is, for his prose treatment of visual phenomena.

As Woolf sees it, the Georgians distinguish themselves through their sight—their eyes are “in the carriage” with Mrs. Brown, rather than lost in utopias, or concerned with grand societal problems. They see Mrs. Brown in detail, but they are not carried away with detail itself, and do not attempt to paint a complete picture. It is no coincidence that “Mrs. Brown”—Woolf’s invented name for this real person—is also the name of a color. Woolf’s description of her lists things that are likely brown, without needing to use the word: she is “threadbare,” and wears “clean little boots” (- Woolf, Collected Essays by Virginia Woolf. Vol. 1 323). Her appearance suggested to Woolf “extreme poverty” without needing “rags or dirt.” Her only line of dialogue is, “can you tell me if an oak tree dies when the leaves have been eaten for two years in succession by caterpillars?”—a scene again full of brown things, like dead trees. In contrast, the villain of this story, Mr. Smith, is considerably richer, sporting “blue serge.”

An even more substantive event of the year is Roger Fry’s exhibit in 1910. The art critic and painter Roger Fry, a late entry into Woolf’s Bloomsbury group of friends, organized, at the Grafton Galleries, one of the most influential British art exhibitions of the early twentieth century: “Manet and the Post-Impressionists,” the exhibition which coined the term “post-impressionist.” Like its predecessor, impressionism, post-impressionism places greater importance on bright color and form than in representation, drama, or chiaroscuro. Its apotheosis, in terms of this emphasis, is fauvism, the trend, circa 1905–1908, for reducing paintings to large swaths of extremely bright colors.

To speak precisely, post-impressionist paintings are, according to my image analysis, 37.4% brighter than European paintings of the early nineteenth century.I arrive at this figure by querying Wikidata for paintings labeled with the movement “post-impressionism,” then querying for paintings created between 1800–1850, downloading one hundred images from each category, and finally by computing K-means centroids for their luminosity in HSL color space (hue, saturation, luminosity). Although we must account for some fading of the paintings’ colors over time, this figure confirms my intuitive sense (and that of many art critics) that post-impressionism was an exceptionally bright artistic movement.

The exhibition flaunts this new brightness, featuring the titular Edouard Manet, along with Paul Cézanne, Paul Gauguin, Maurice Denis, Vincent van Gogh, and other painters known for their use of bright color. It was highly controversial. A cartoon in the November 1910 issue of The Bystander, as shown in fig. 2, illustrates the controversy (The Bystander 375). A segment of the cartoon labeled “you arrive thus” depicts two gentlemen arriving at the exhibition: one half-asleep, and the other, alert. The following segment, labeled “and depart thus” depicts the sleepy man suddenly alert, and the alert man with buckled knees, wiping the sweat from his brow. Another corner of the cartoon depicts an appreciative look from “an American art student who liked the colour harshness.” Finally, a tableau at the bottom of the page shows four men and three women, in genteel hats and costume, doubled over in laughter. The caption is “From the Pictures’ Point of View.”

Even Roger Fry’s own comment on his exhibition had, in his biographer Virginia Woolf’s phrase, “an apologetic air” (Woolf, Roger Fry 153). “There is no denying,” Woolf quotes Fry as saying, “that the work of the Post-Impressionists is sufficiently disconcerting. It may even appear ridiculous to those who do not recall the fact that a good rocking-horse has often more of the true horse about it than an instantaneous photograph of a Derby winner.” Fry contends that an abstracted, artificial reproduction, due to its artificiality, is, paradoxically, more real than its realistic depiction. This introduces a theme which we shall see several more times: that raw colors, abstractions, and blocky, Platonic shapes, like that of a rocking horse, are somehow more faithful to the phenomenological aesthetic experience than that of ostensibly objective representation, such as photography.

The reactions to this exhibition were not an exaggeration. As Woolf put it:

“it is difficult in 1939, when a great hospital is benefiting from a centenary exhibition of Cézanne’s works, and the gallery is daily crowded with devout and submissive worshippers, to realize what violent emotions those pictures excited less than thirty years ago. … The public in 1910 was thrown into paroxysms of rage and laughter. … they were infuriated. The pictures were a joke, and a joke at their expense. … The pictures were outrageous, anarchistic and childish. They were an insult to the British public and the man who was responsible for the insult was either a fool, an impostor or a knave” (Woolf, Roger Fry 153–54).

“Anarchistic and childish.” These reactions, while extreme, bear some examination. First, “childish” is telling, given the well-documented preference among children for bright colors (Boyatzis and Varghese; Nanda et al.). In fact, Woolf relates the experience of fellow Bloomsbury member Desmond MacCarthy, to whom “parents sent … childish scribbles which they asserted were far superior to the works of Cézanne” (Woolf, Roger Fry 154). Second, the reaction that the paintings were “anarchistic” speaks to a related thread of so-called primitive art that was incipient in modernism. These are observations which art critic C.J. Holmes also makes of the exhibition:

The tradition of Post-Impressionism, then … is the expression of a personal vision: 1. Through the methods, first applied to oil-paintings by the Impressionists, which aim at the greatest possible vibrancy and luminosity of colour, obtained by the juxtaposition of pure bright pigment in small separate touches. … 2. Through rigid simplification on the lines of the Orientals and of Daumier, in which the means of expression are reduced to line and colour … (Holmes 19)

The “pure bright pigment in small separate touches” is a technique borrowed from impressionism, and its sister style, pointillism: it is one which leverages the ocular and neurological interactions of color to produce new colors.See the section on impressionism below for a more thorough treatment of color theory in impressionism, and its relation to literary impressionism.

The “lines of the Orientals” speaks to the influence of far Eastern art—in particular Japanese woodblock prints—that would become important for modernism, and which would reach to its climax in the 1920s, with such works as “The Waste Land” in 1922, and the first British performance of Stravinsky’s ballet “The Rite of Spring” in 1921. As we will see later, rich visual descriptions are often correlated with geographic, temporal, or imaginitive distance, and so appear often in travel narratives, period pieces, science fiction, and other genres of distance.

But the point I am making here is not only one of the art-historical milieu in which modernist literature is steeped, but that this way of thinking about color perception is crucial to understanding the use of color words, in descriptive prose or poetry, since color words, like “pure bright pigment,” are single brushstrokes that belong to their greater pictures—that is, their textual context.

The influence of this exhibit, for the writers, thinkers, and artists living in London, is difficult to overestimate. This is partially owed to its colorfulness: it was one of the first art exhibitions in Britain to be about color. Narrative, and to a lesser extent representation, play lesser roles in these paintings than does color. Although the impressionists of the prior decades always had noticeably bright color-palettes, and were more interested in the interplay of color than anything else, the post-impressionists of this exhibit attacked their viewers with color. This is the climate in which the colors of modernist literature were written.

A third significant event of the 1910s was the arrival, in the early part of the year, of Halley’s Comet. A literal once-in-a-lifetime event, happening only every 75 years, the comet had long appeared as an omen at various turning-points in British and world history. It appeared in the year of the Norman conquest of England in 1066, and was depicted in the Bayeux Tapestry, which is its first known representation. In the late seventeenth century, British astronomer Edmond Halley, in dialogue with Isaac Newton, noted its periodicity. Its arrival in 1910 was special, however, since it would come closer to the earth than ever before. Furthermore, recent developments in spectrometry found, in 1909, that its tail contained the poisonous gas cyanogen (Stoyan 147). This caused mass hysteria in Europe, where gas masks, bottles of oxygen, and “comet pills” were sold in great quantities.

Comet-panic was fueled by comet-related catastrophe in science fiction: Edgar Allen Poe’s story “The Conversation of Eiros and Charmion,” Jules Verne’s novel Hector Servadac (or “Off on a Comet,” in English editions), Camille Flammarion’s La Fin du Monde, and H.G. Wells’s 1906 In the Days of the Comet, the only one to speculate about positive effects of extraterrestrial gases. There, the comet’s gases have the effect of opening the minds of the world’s denizens, allowing them to engage in free love and polyamory. In all of these narratives, however, comets are not mere omens, “shooting stars” in the otherwise static firmament, existing only as signals for astrologers, but are informed by the astronomical knowledge that they are large physical objects which may crash into Earth.

Although British intellectuals, on the whole, did not buy in to this comet panic, or liberation, its effects were still felt. John Maynard Keynes, an economist in the Bloomsbury Group, wrote, “the comet’s tail last night produced a very odd state of affairs in the air here. It was oppressive & electric & rather exciting” (Stansky 120).

George Dangerfield’s influential 1935 history, The Strange Death of Liberal England, begins with the 1910 appearance of Halley’s Comet, and the death of King Edward VII. “Upon the chill and vacant twilight blazed Halley’s Comet – which, visiting the European heavens but once in a century, had arrived with appalling promptness to blaze forth the death of a king” (Dangerfield 19). He writes, in his characteristically dramatic style, that 1910 is “a landmark in English history, which stands out against a peculiar background of flame. For it was in 1910 that fires long smouldering in the English spirit suddenly flared up, so that by the end of 1913 Liberal England was reduced to ashes” (15-16).

These three events are reflective of, if not emblematic of, the changes in human character which Woolf identifies as happening in 1910: they signal the strange death of the old beige order, and the coming of a new, brightly colored era. These are phenomena we see in the results of my computational analysis below. But to understand the wave of color in 1910, we need to go back a little further, to the 1890s.

The Yellow Nineties: Decadence, Cosmetics, and Artificiality

Despite what Woolf would have us believe, human character, and with it a sense of colorfulness and brightness, didn’t change all of a sudden in 1910. Rather, there was a gradual change, with a large number of causes. Even the 1910 post-impressionism exhibit had its origins a decade or more earlier, in France. One of the most significant of these cultural roots, and one to which Woolf alludes, is the environment of the late Victorian period: the 1890s, or “yellow nineties” as it was often known. This period saw a trend towards the celebration of artificiality: synthetic pigments and bright colors, exemplified by artistic movements such as the aestheticists and the pre-Raphaelites. Yellow was its color in more than one respect. As Frances Winwar puts it:

The color had been a favorite one with Rosetti, Morris and Burne-Jones who had first discovered it in the richness of medieval panels, then nearer at hand in the sunflower of rural gardens. … There was something vivid and daring about the color, something of the times, like the golden bloom of the age on a century that was nearing its close—not to death but to greater achievement. Yellow and fin-de-siècle began to have connotations open to many meanings but all leading to a definable sense of modernity, challenge, emancipation. People were no longer afraid to live. … The public became yellow-conscious. (Winwar 239)

Yellow had become a symbol of modernity. It was a bright color, comparatively rare in nature—the color of the sun, of gold, and of sunflowers. It became the symbol of aestheticism. Fig. 3 shows the proportions of the hypernym yellow, according to the publication dates of the works where it appears. This isn’t just the word yellow, I should reiterate, but its hyponyms, associated lemmas, and objects that are typically yellow-colored: lemons, sunflowers, and so on. Although proportions of textual color generally increase with time, in literature of this period, yellow was never higher than it was in the 1890s.

Richard le Gallienne, English poet and essayist, and a prolific writer throughout the nineties, described in his 1896 Prose Fancies what he called “The Boom in Yellow,” in which colors like green and white had fallen out of fashion, to be replaced with yellow. This essay makes his Prose Fancies the work with the highest proportion of yellow in the corpus:

Innocence has but two colours, white or green. But Becky Sharp’s eyes also were green, and the green of the aesthete does not suggest innocence. There will always be wearers of the green carnation; but the popular vogue which green has enjoyed for the last ten of fifteen years is probably passing. … now the triumph of yellow is imminent. Of course, a love for green implies some regard for yellow, and in our so-called aesthetic renaissance the sunflower went before the green carnation—which is, indeed, the badge of but a small schism of aesthetes, and not worn by the great body of the more catholic lovers of beauty. (Gallienne 85–86).

Le Gallienne refers to the green carnation, a frequent lapel adornment of the London aesthetes in the nineties. As carnations are not naturally green, but must be dyed to achieve this color, using a blue-green aniline dye,A synthetic dye. See the section on materialities below.

this symbolized the artificiality prized by aestheticism, namely, unnaturally bright colors. As Joseph Valente and others have noted, they were “an aestheticist emblem of imaginative artifice … and a badge of the homosexual subculture of fin-de-siècle England” (Valente 251). Oscar Wilde reportedly bade his entourage wear green carnations at the opening of Lady Windermere’s Fan, in order to “annoy the public” (Sturgis 423). When asked why he wanted to annoy the public, Wilde replied, “it likes to be annoyed.”

The Green Carnation is also the title of Robert Hichens’s 1894 roman à clef, first published anonymously, which satirizes Wilde, his relationship with Lord Alfred Douglas (not yet then public), and his coterie. Although it uses colorful description in a parodic mode, its parody is achieved by overemphasizing color. Here is a passage from its opening pages, annotated with the colors inferred from the computational color model I will soon introduce:

Floating through the curtained window the soft lemon light sparkled on the silver backs of the brushes that lay on the toilet-table, on the dressing-gown of spun silk that hung from a hook behind the door, on the great mass of gloire de Dijon roses, that dreamed in an ivory- white bowl set on the writing-table of ruddy -brown wood. It caught the gilt of the boy's fair hair and turned it into brightest gold, until, despite the white weariness of his face, the pale fretfulness of his eyes, he looked like some angel in a church window designed by Burne-Jones...

(Hichens)

There are some unusual color descriptors here: “soft lemon,” “gilt,” “brightest gold.” And where the colors are more conventional, the objects they describe are unusual: “the gilt” refers to the boy’s hair; “white” to his “weariness,” “pale” to his “fretfulness.” This is the playful mood out of which the “yellow boom” was born.

Le Gallienne knows this kind of yellow light, which streams into the room, for in the “The Boom in Yellow,” he observes that “when the sun comes out upon a yellow wall-paper the whole room seems suddenly to expand, to open like a flower” (Gallienne 87). Incidentally, the famous short story by Charlotte Perkins Gillman, “The Yellow Wallpaper,” had just been published, in 1892. In 1894, a novel appeared by the psydonymous “Iota,” called A Yellow Aster, the popularity of which Le Gallienne cites as an indicator of the “boom.” And The Yellow Book, the literary and arts journal to which Le Gallienne frequently contributed, began to be published, in the same year. Le Gallienne argues that, “with any other colour, it would hardly have sold as well” (Gallienne 87).

The Yellow Book was a short-lived illustrated quarterly journal of literature and art, published between 1894–1897. The bright yellow color of its cover recalls the look of racy French novels of the day, which were similarly bound (Wilde xx). It was also the color of avant-garde literature, and edgy books of all types (Doran 3). Priced at five shillings, it was comparatively cheap, as well, costing less than a new volume of fiction (Brooker and Thacker 78). Fig. 4 shows the cover of the first issue.

Holbrook Jackson describes its impact thus: “Nothing like The Yellow Book had been seen before. It was newness in excelsis: novelty naked and unashamed. People were puzzled and shocked and delighted, and yellow became the colour of the hour, the symbol of the time-spirit. It was associated with all that was bizarre and queer in art and life, with all that was outrageously modern” (Jackson 54).

Its inaugural issue began with a sketch by the pre-Raphaelite painter Frederich Leighton, the Henry James story “The Death of the Lion,” which parodies literary canonization, and an essay by Max Beerbohm titled “A Defense of Cosmetics.” Beerbohm’s essay begins: “it is useless to protest. Artifice must queen it once more in the town” (Beardsley 65). The “artifice” he exhalts is that of cosmetics, or the cosmetic, more generally. The epoch, as he sees it, is changing, and this has a distinctly visual effect: “For behold! The Victorian era comes to its end and the day of sancta simplicitas is quite ended. The old signs are here and the portents to warn the seer of life that we are ripe for a new epoch of artifice. Are not the men rattling the dice-box and ladies dipping their fingers in the rouge-pots?” (Beardsley 65). Rouge, of course, is the cosmetic product, usually worn on the lips and face, and which takes its name from the French word for red. It is usually reddish in color, but not always. It would not have been fashionable for Victorian women, apart from actresses, to use rouge, since, as one historian of cosmetics put it: “women of the [Victorian era] had to disguise any attempts at self-improvement. The prudery of contemporary moral standards was totally prohibitive as far as female vanity was concerned …” (Gunn 137). But this was starting to change by the 1890s.

For Beerbohm, and many other contributors to The Yellow Book, modernity meant bright colors: yellow books, yellow paintings, and red lipstick. “The era of rouge is upon us,” he wrote. “The good combinations of line and colour are nearly numberless … No monotony will be” (Beardsley 78).

The Yellow Book was, depending on who you talk to, either a Victorian precursor to the modernist little magazine, or one of its first exemplars. Either way, there’s no question that “the boom in yellow” became a boom in all colors, shortly after the ’90s came to a close.

Materialities of Color at the Turn of the century

Much of what explains the explosion of color, at the end of the nineteenth century, was material: the development of industrial pigments. There were simply more colors available in everyday life. Clothes could be dyed much more cheaply, using synthetic dyes; so could commercial products of all sorts. Paints, interior and exterior, could be made which were brighter, longer-lasting, and representing a wider spectrum. Advances in color printing techniques, such chromolithography and chromo-photo-lithography, made publications much brighter than before. As Blaszczyk and Spiekermann put it in their monograph of commercial color, Bright Modernity: “the mills that produced silk, cotton, wool, and rayon fabrics for use in ladies’ dresses and men’s suits relished brilliant hues that did not fade in the sun and blacks that were truly black. Retail shops created eye-catching window displays using electric lights and brightly colored fabric backdrops to attract window shoppers after dark” (Blaszczyk and Spiekermann 2).

We owe the expansion of color words in part to the industrial revolution, and to the marketing of textiles and other goods with colors. One famous example is that of mauve, a very rare word until around 1860, when the dye mauvine began to be used commercially (“Mauve, Adj.”). Previously, one of the most expensive pigments worldwide was Tyrian Purple (#66023C), and clothes dyed with it were associated with royalty, who were usually the only ones who could afford it. (Although the hue could be approximated from mixtures of other dyes, it would never be as brilliant or as long-lasting as when it was made from its traditional source: a sercretion from rare sea snails.) That changed in 1856, when William Henry Perkin accidentally synthesized from coal tar a substance he would call mauveine, the first aniline dye (Finlay 354). While this may seem like a footnote in industrial chemistry, Perkin was hailed as a folk-hero: upon his arrival in New York, the New York Herald ran an article with the headline: “Coal Tar Wizard, Just Arrived in Country, Transmuted Liquid Dross to Gold” (Garfield 4). Once mauvine entered mass production, the color mauve took on such popularity that the 1860s were termed the mauve decade.Strangely, this is not the mauve decade in America, as Thomas Beer terms it in the literary-social history The Mauve Decade, which chiefly deals with the 1890s (Beer).

The color word mauve tracks the popularity of the dye. A Google Ngrams search of the English Fiction corpus for mauve, shown in fig. 5 shows a dramatic increase of the word’s frequency after 1860. After 1870, however, the word (as, presumably, the color) falls into disfavor from which it won’t recover until the twentieth century.

Incidentally, the data journalist Ben Blatt, in a book of computational literary text analysis descriptively titled Nabokov’s Favorite Word is Mauve has identified mauve as Vladimir Nabokov’s most frequent “cinnamon word,” that is, the word in his writing which is statistically much higher than it is in the Corpus of Historical American English (Blatt 170). Blatt suggests that this may have something to do with the Russian-born novelist’s synesthesia: in his autobiography Speak, Memory, he sees “projected, as it were, upon the inside of the eyelid … a mauve remoteness melting beyond moving masts” (Quoted in Blatt 170).

Mauveine soon paved the way for the synthesis of other aniline pigments, derived from coal tar. Notably, fuchsine was invented that same year, the substance to which we owe the color term fuchsia. Like mauve, both fuchsia (#ED0DD9) and mauve (#AE7181) refer to flowers with the same names (mauve is French for mallow), and so while the words appear earlier than 1856 in the OED, according to Google Books data, the words are almost never used as colors. In fact, their primary definitions refer to color of the aniline dye, specifically (“Mauve, Adj.”). The words mauve and fuchsia describe not just colors, but artificial colors, as does Beerbohm’s rouge. The number of artificial dyes would grow from fewer than 50, in 1870, to 1300, by 1913 (Blaszczyk and Spiekermann 3).

With this explosion in the availability of pigments came a need to name all the new colors, and to construct a scientific language of color which could be used to ensure color consistency in industry. Thus the genre of color manual was invented. They were used by naturalists, in need of describing the colors of new species; by philatelists, in need of describing the color of their stamps, and by manufacturers, who needed to keep a consistent look across disconnected factories. They were also used by educators, most famously by American educational reformers Milton Bradley and Albert Munsell, who developed systems of color education which were inspired by researchers such as Elizabeth Peabody and Friedrich Froebel, the inventor of kindergarten. Bradley and Munsell saw the need to standardize color nomenclature for childhood education, and so published books of color samples which could be used as universal referents. It is these books, we shall see in the section on heuristic color maps below, which I will use as inputs to train a model of literary imagination.

First, however, we must fully understand the problems presented by modeling visuality, by closely examining color language in two modernist exemplars: Virginia Wool’fs To The Lighthouse and James Joyce’s A Portrait of the Artist as a Young Man.

Case Study: Colors in To the Lighthouse

The works of Virginia Woolf make an exemplary case-study for a analysis of color words, not only because Woolf wrote so eloquently in her prose about the brightening colors of her generation, as I discuss above, or even that her works rank among the highest in terms of raw color proportions, as we will see in the results below in table 6, but because her fiction is remarkably painterly: at times impressionistic, at times cubistic, but always deeply visual. In Woolf’s own term, as I will explain below, it is all eye.

Since the Bloomsbury Group was so strongly intermedial—its members not only writers, but artists, theorists, biographers, critics, curators, and more—a tremendous amount has been written about the interplay between these people and their respective media. Jane Goldman’s The Feminist Aesthetics of Virginia Woolf reads Virginia Woolf’s aesthetic practices through her feminism, especially where it concerns post-impressionism. Goldman discusses the Woolfian moment which seems to stretch time, as in Bergson’s durée, such that sensory perception is heightened, colors brightened (Goldman). Claudia Olk’s Virginia Woolf and the Aesthetics of Vision takes a more philosophical approach, studying vision as “a significant semantic and structuring principle of Woolf’s novels as well as one of their organising paradoxes,” and putting Woolf into dialogue with writing on vision from Plato, Heidegger, and others (Olk 17). The Multiple Muses of Virginia Woolf collects a range of essays, connecting Woolf with Proust, Fry, 1920s cinema, Russian ballet, and music; and Virginia Woolf’s Bloomsbury collects essays on aesthetic theories within the Bloomsbury group (D. Gillespie; Potts). Maggie Humm’s Modernist Women and Visual Cultures examines Virginia Woolf and Vanessa Bell’s relations with photography and cinema, and publishes many family photos taken by the sisters (Humm). Jack Stewart’s Color, Space, and Creativity, cited earlier, is an analysis of the work of five British writers, the first of which is Woolf (Jack Stewart 25). His analysis hinges on cataloguing symbolisms of sorts: “various reds form a masculine complex consisting of Mr. Ramsay’s red-hot pokers, red geraniums, and reddish brown hedge; the reddish brown stocking that Mrs. Ramsay knits for the lighthouse keeper’s son; he image of James in judge’s robes; Paul Rayley’s blaze of amorous passion; and Charles Tansley’s red raucousness” (Jack Stewart 28).

The relationship between Woolf and her sister has been long discussed, as well. Both artists, a writer and a painter, they are at once a classic case of sibling rivalry, as well as a reification of the sister arts of painting and writing. Much has been made of the way they would spend time together, with Virginia writing or reading, and Vanessa painting. Or, how they would respond to one another’s work, with Woolf writing prefaces to Bell’s exhibition catalogs, and Bell designing covers for Woolf’s books. Diane Gillespie’s The Sisters’ Arts deals with just this: the sisters’ collaborations, critiques, and “dual creativities” (D. F. Gillespie). Jane Dunn’s A Very Close Conspiracy takes a more biographical approach to the sisters’ symbiosis (Dunn). And at least two novels have been written which fictionalize their relationship (Parmar; Sellers).

But what I would like to focus on, just by way of introducing this case-study, are the ocular and chromato-phenomenological aspects of Woolf’s aesthetics. That her work is all eye is not just my metaphor. She often wrestles with the interplay of vision and text in her own prose writing. For instance, color is the central topic of conversation in her 1934 prose work, Walter Sickert: a Conversation. The slim volume of only 28 pages begins at a dinner party, where one of Woolf’s guests claims that “in the eyes of the motorist,” “red is not a color but simply a danger signal” (Woolf, Walter Sickert 5). Another guest adds that, due to the increasing proliferation of colored signals, “we shall very soon lose our sense of color.” As the conversation evolves, they discuss many of the aspects of color I discuss in this chapter:

…how different people see colour differently; how painters are affected by their place of birth, whether in the blue South or the grey North; how colour blazes, unrelated to any object, in the eyes of children; how politicians and business men are blind, days spend in an office leading to atrophy of the eye; and so, by contrast, to those insects, said still to be found in the primeval forests of South America, in whom the eye is so developed that they are all eye, the body a tuft of feather, serving merely to connect the two great chambers of vision. (Woolf, Walter Sickert 7)

Woolf juxtaposes the pragmatism of “politicians and business men” with children and insects: beings for whom the sense of sight, and the sensation of color, are greatly magnified. This reminds another guest of a recent outing to see paintings by the British post-impressionist Walter Sickert:

…When I first went into Sickert’s show, said one of the diners, I became completely and solely an insect—all eye. I flew from colour to colour, from red to blue, from yellow to green. Colours went spirally through my body lighting a flare as if a rocket fell through the night and lit up greens and browns, grass and trees, and there in the grass a white bird. Colour warmed, thrilled, chafed, burnt, soothed, fed and finally exhausted me." (Woolf, Walter Sickert 9)

Many of Sickert’s works contain exceptionally bright colors, straight from the tube, they often seem. His choice of subject matter often tells a story, not unlike those of a novelist. Like Edward Hopper, and other painters of the decades to follow, the luminosities of his paintings contrast sharply with a subject matter that can portray deep loneliness, ennui, regret, or other subtle forms of discontent. In fact, these are often the titles of his paintings: see his Ennui (1913) in fig. 6, Despair (1908), or What Shall We Do for the Rent?.

Woolf’s dinner guests therefore describe him as a “biographer,” painting portraits-as-biography, or a novelist (Woolf, Walter Sickert 13). But the important question, for them, is “to what school of novelists does he belong?” (Woolf, Walter Sickert 17). “He is a realist,” one answers, “… nearer to Dickens than to Meredith. He has something in common with Balzac, Gissing and the earlier Arnold Bennett” (Woolf, Walter Sickert 17). The conversation describes paintings as if they were prose, and prose as if they were paintings—if not synaesthetically, then at least intermedially. One guest remembers reading a letter from Sickert in which he says “I have always been a literary painter, thank goodness, like all the decent painters” (Woolf, Walter Sickert 26).

But again, image-in-prose is never as simple as the Horatian ut pictura poesis. Color in text and speech always derives its effect through collocations, recollections, and other associations. As one dinner guest puts it:

[Sickert] must often think that to describe a scene is the worst way to show it. It must be done with one word, or with one word in skilful contrast with another. For example, there is Shakespeare’s ‘Dear as the ruddy drops that visit this sad heart.’ Does not ‘ruddy’ shine out partly because ‘sad’ comes after it; does not ‘sad’ convey to us a double sense of the gloom of the mind and the dullness of colour? They both speak at once, striking two notes to make one chord, stimulating the eye of the mind and of the body. (Woolf, Walter Sickert 22–23)

This guest compares this to the bright colors of a petticoat and chest of drawers: “…when, for example, we said that Rose’s red petticoat satisfied us; … Why did the red petticoat, the yellow chest of drawers, make us feel something that had nothing to do with the story? We could not say; we could not express in words the effect of those combinations of line and color” (Woolf, Walter Sickert 25).

These inexpressible color-sensations, what philosophers might call inverted qualia(See, for example, Byrne; ).

, are features of Woolf’s fiction, as well as her nonfiction. To the Lighthouse is a perfect text for examining the role of color in literature, since not only is it very colorful, but deals explicitly with color perception. This is a novel that treats raw color perception, and raw sensation—as opposed to conventionally linguistic color writing—as transformative.

On one of the first pages, we hear Mrs. Ramsay muse that “any turn in the wheel of sensation has the power to crystallise and transfix the moment upon which its gloom or radiance rests” (Woolf, To the Lighthouse 3). On the one hand, as a novel that deals with an artist and painting, this is an extreme example of the phenomenon I’m outlining in this chapter, but on the other, this is also the best example. Let’s start with how the novel treats painting, and its relation to its subject:

There are many phenomena of note in this colorful passage. Although we are not in painter Lily Briscoe’s mind, in this descriptive narration, we nonetheless see this scene painted with many simple, primary colors: unmixed colors, straight from the tube. First, the water is so flat, or so blue, as to resemble a plate. Then, the green grasses, and the implied sand-color of the dunes. So far, this seems like a stereotypical seaside landscape. But even though this is a tableau, constructed precisely to resemble a painting, it is not at all static: the dunes are “fading and falling,” the grasses are “flowing,” and “running away.” Both the lighthouse and the grasses are anthropomorphized, to some degree. It is only when the painter Mr. Paunceforte approaches it that it is reduced to simple “grey,” “lemon-colour,” and “pink”: subdued, secondary colors. These are the colors against which Lily Briscoe rebels, in her own painting of these scenes. Here is Lily’s free indirect discourse:

In contrast to Mr. Paunceforte, who paints to create a pleasant, “elegant” painting, Lily pledges fidelity to the bright violet of the jacmanna, and thus to her own perception of the color, no matter how inelegant it might seem. We see these same colors appear moments earlier, when Lily, “with all her senses quickened as they were,” was “looking, straining, till the colour of the wall and the jacmanna beyond burnt into her eyes.” Lily is so devoted to faithfully conveying the color of this scene that she has allowed her eyes to unfocus, and her vision to blur, impressionistically, which softens the edges of the scene, and reduces it to just its colors.

Here again, textual colors are the vectors along which visual associations take place: transitions from one thought to the next. They enact a persistence of vision in prose. When Mrs. Ramsay imagines her son “all red and ermine on the Bench,”The court dress of Lord Justice Clerks, among other judges, is red and ermine.

that color is repeated in, or prompted by, the reddish-brown stockings that she knits for her son, only a paragraph later. Woolf suggests, then, through this chromatic association, that she knits him these stockings as an unconscious way of preparing him for a future career that she imagines for him. But the key is that she imagines him red, not his clothes, suggesting that this color impressionistically overtakes the image. It is a blur, a composite image, as in a dream.

When we examine the incidence of colors along the narrative time of the novel, as in fig. 7 with the x-axis representing ten sections from the novel’s beginning to its end, we see an overview of its narrative-descriptive arc.

The parts of the novel with the most color are undoubtedly the beginning and the end. But a close contender is the middle section 7, which, as readers of this novel have no doubt already guessed, aligns perfectly with the “Time Passes” section. This is a strikingly poetic segment of the novel, full of abstract language, nature imagery, and few people. As in poems, and poetic description, time is allowed to run wild: narrative, plot, and character become subservient to vision and perception. Again, although these images at times painting-like perceptions, they are extremely dynamic, as if a film is being played at four times its recorded speed.

The novel ends just as colorfully as it started, and with an appropriate image. Lily Briscoe finishes her painting, and looks at her canvas: “it was blurred,” we are told. Finally, Lily says, “I have had my vision.” Vision, as I have argued in the introduction to this dissertation, is a curious word, since it is almost always used metaphorically. The times we encounter a phrase like 20/20 vision in literature of this period are far outnumbered by the times we see vision used in the sense of imagination, prediction, plan, or clairvoyance, as in, for instance, Yeats’s A Vision, or H.D.’s Notes on Thought and Vision (Yeats et al.; Doolittle). These are all, paradoxically, modes of thinking, or of intuition, that don’t involve actual sight. But yet the superficial meaning of this term is that of seeing. This is more than a chance ambiguity, but a testament to the discourse between visual experience and thought.

Case Study: Colors in A Portrait of the Artist as a Young Man

Another canonical modernist work with an abundance of textual color is James Joyce’s autobiographical bildungsroman, A Portrait of the Artist as a Young Man. Portraits are a genre of painting that are mimetic by definition: they aim to represent the image of their subjects. But the portraits of modernist Bildungsromane, however autobiographical, are impressionistic, representing instead the unfiltered worlds of sensory impressions, in order to mimic the visual experience of children, who have not yet mastered pragmatic lexical categorizations of their visual worlds. Bildungsromane often mirror this developmental phenomenology: being novels of education, they usually proceed chronologically, beginning with youth, and often from the viewpoint of youth. Joyce’s A Portrait of the Artist as a Young Man is the classic example of this narrative style.

Compared to Woolf and To the Lighthouse, Portrait has comparatively little critical writing which addresses his use of colors. Although a few studies directly address the use of color in Portrait. John O’Sullivan’s monograph Joyce’s Use of Colors chiefly deals with Finnegans Wake and Ulysses, but in it, he comments that “the color patterns of A Portrait of the Artist as a Young Man are most associated with the ambiguities present in the novel, especially with the confusion and misunderstandings of the protagonist” (Sullivan 23). Yutong Xie’s article “Color as Metaphor” examines Joyce’s use of black and green in Dubliners and Portrait (Xie).

Elizabeth Switaj’s James Joyce’s Teaching Life and Methods reads Portrait through Joyce’s pedagogical practices, and has a chapter on Portrait (Switaj). Switaj explains the color terminology used in Portrait through the lens of the Berlitz method, from which Joyce was teaching, in his day-job as an English teacher. In Method for Teaching Modern Languages, for example, the Berlitz Textbook he used, colors are on the very first page of instructional material. “Colours:” it reads, “red, blue, yellow, green, black, white, gray, brown. The pencil is green, the book is blue, the ruler is yellow, the necktie is red, the boot is black, the coat is gray, the hat is brown” (Berlitz 18). (Note: orange and violet are conspicuously absent. What would one of Joyce’s students have said, had they been looking at an orange book, or a purple coat?) Switaj argues that Stephen’s vocabulary of color words, like much of the rest of his vocabulary, follows the vocabulary of the Berlitz textbook (Switaj 52).

David Kastan, a professor of English who collaborates with a visual artist in the treatise On Color, begins with the first page of Portrait, however, the page of which features young Stephen’s refracted sensory impressions (Kastan 2). Kastan notes that young Stephen’s observation that “you could not have a green rose” is predicated on the aphorism that “roses are red.”

The first page of Portrait is what Derek Attridge calls “one of the most revolutionary pages in the history of fiction” (Attridge). Among them is his lisped, misremembered version of his father’s apparent bowdlerization of H.S. Thompson’s song, “Lily Dale”:

Steven’s father, Simon Dedalus, substitutes “place” for what is in Thompson’s song, “grave,” presumably to make the song friendlier to his son. The song’s chorus actually reads:

This mistake is the first of many parapraxes which will become hallmarks of Joyce’s style. Young Stephen’s refraction, “o, the geen wothe botheth,” which makes the rose green, rather than the grave or place, appears several pages later, when an older Stephen muses in the classroom at Clongowes, where the class has been divided into white and red rose factions, an imitation of the English War of the Roses.

Green roses are of course possible with artificial coloring, as with green carnations, Wilde’s aestheticist symbol that I’ve discussed above. This has led some to suggest that the “wild” of the “wild rose” may be a pun for “Wilde” (Valente 251).

But in the geopolitical climate of Ireland of 1916—the year Portrait was published, and the year of the Easter Rising—green roses are hardly aesthetic-as-apolitical, that is, art-for-art’s-sake. Just as white and red roses symbolized the warring houses of Lancaster and York, green was a nationalist color for Irish homerule and independence. So when H. G. Wells’s 1917 review of Portrait celebrates the novel for its “too true” “account of the political atmosphere in which a number of brilliant Irishman have grown up,” an atmosphere he diagnoses as “just hate, a cant cultivated to the pitch of monomania,” his epithet for these young Irishmen is “bright-green”: colored by nationalism, but also naive, inexperienced (Wells, “James Joyce” 88).

Katherine Mullin notes that the heightened sensory detail of young Stephen’s narration “shows the interior thoughts of a character expressed in the language he might use at the time his thoughts are occurring,” a technique she identifies as “coloured narrative” (Mullin). The term is Graham Hough’s, from a 1970 essay on Jane Austen, and refers to a narrative mode akin to free indirect discourse (Hough 205). The color of the term is meant metaphorically, and does not refer to literal colors, but is nonetheless an apt metaphor for Stephen’s colorful prose: his narration is not only colored by his personality, as Hough hears in the narrative of Emma, but by the visual qualities of his heightened perception.

Portrait, in its colored narrative, exhibits the modernist struggle with subjectivity through color, the locus of a great problem with knowledge and objectivity.

Fig. 8 shows the proportions of color in Portrait, plotted in narrative time. There are four main areas of color concentration here. The beginning, where Stephen is young and full of strong sense-impressions, exhibits the highest concentration. The next is at the halfway point of the novel (point 5). This is a segment from Father Arnall’s hell sermon, a winding, verbose speech about the horrors of hell, here annotated with \(CM_X\):

This segment is notable, not because it uses traditionally-recognizable abstract color words, like red, green, blue, and so on, but because it is colorful for its imaginable nouns: sand (#e2ca76), earth (#a2653e), forest (#0b5509), ocean (#017b92), sky (#82cafc), and sea (#3c9992), for example. It is no coincidence that Father Arnall repeats, three times, his instruction for his young listeners to imagine.

The next peak in colors comes at 60% of the novel, where an older Stephen, an aspiring poet, wonders what it is that attracts him to words. He reads one of his lines, “a day of dappled seaborne clouds,” and all of a sudden, “the phrase and the day and the scene harmonized in a chord” (Joyce, Portrait 146). Stephen thinks:

But Stephen soon negates himself: it is not the colors of words, or color words, that enchants him:

No, it was not their colours: it was the poise and balance of the period itself. Did he then love the rhythmic rise and fall of words better than their associations of legend and colour? Or was it that, being as weak of sight as he was shy of mind, he drew less pleasure from the reflection of the glowing sensible world through the prism of a language many-coloured and richly storied than from the contemplation of an inner world of individual emotions mirrored perfectly in a lucid supple periodic prose? He passed from the trembling bridge on to firm land again.

He never answers this question, which seems to linger on “the trembling bridge” of language, a bridge which spans, on the one hand, vision, imagination, and color: the concrete, the plastic; and on the other hand, linguistic music: the sounds and rhythms of words. Stephen’s phrase, “a day of dappled seaborne clouds,” is not just an image, containing cloud and sea, and recalling the “dappled things” for which Gerard Manley Hopkins praises God in “Pied Beauty,”Although written in 1877, this poem would not be published until 1918, making it too late to have been an influence for Joyce. Had Hopkins live a little longer, or had Joyce been born a little later, they surely would have met: a Jesuit, Hopkins taught at University College Dublin, which Jesuit-educated Joyce attended.

but music: four alliterative iambs.

Like Woolf, and so many other modernist writers, Joyce describes literary language using chromatic metaphors, because color, in text, is evocative of mental images. The “sensible world” “reflects” through the “prism” of language, a device which, as Newton demonstrated, splits its input into a spectrum. This prism is that of imagination: the writer’s, and the reader’s: a mental image is split through the prism of language into its constituent colors, and then reconstituted in the mind of the reader.

The jump in textual color at the end of Portrait reflects the colorful descriptions in Stephen’s journal entries, which conclude the novel. Here is one entry:

Here we have many of the properties of textual color converging. As I’ve discussed above, textual color tends to correlate with high emotions: ecstasy, love, rapture (“O life!”). And color words are color exceptions: they describe things which are visually unusual, or striking: such is the mood of springtime. The first days of spring—in Ireland, at least—are full of new colors: where there had been gray skies and bare trees, there are now brightly-colored apple flowers. People leave their houses to spend time outdoors, “romping.” Stephen stares at the girls, notices their hair, and imagines them blushing (because of something he said?). As Sinclair wrote of Richardson, “it is as if no other writers had ever used their senses so purely.”

Color Epistemology

Before we can discuss the modeling of color information, we must first consider some of the problems with color, with color words, and with the connections between the two. Consider the color term lemon-yellow. A metaphor derived from a natural object, it ostensibly suggests a color which is the yellow of lemons. Unlike abstract colors like red or blue, there is a natural referent to which we all can turn if we want a common understanding of a color. But lemons, as a farmer might explain, are only yellow during a short window of their existence. To further complicate matters, even ripe lemons at a grocery store vary considerably between batches, individuals, and levels of freshness. But at the same time, lemon-yellow is not strictly a Platonic ideal to which all lemons, or paintings of lemons, aspire, either. Instead, the term describes a color phenomenon somewhere between lemons, our memory of them, our visual experience of them, and what we read about them. This semantic slipperiness is a problem to which I’ll return often in this chapter, since it represents what I see as one of the central concerns of modernist literature: gaps in the phenomenology of visual experience, between the writer’s visual experience, the text as written, and the reader’s imagination.

Aloys Maerz and Morris Paul’s 1930 reference manual A Dictionary of Color, one of the inputs to the model I’ll make below, acknowledges this problem as one they hope to solve with their manual. They see this as a part of the “material” and “intellectual” confusions of color names:

The confused ideas on color nomenclature are found due to two factors, one material, the other intellectual. The first has been the ability of color makers, in the past, to produce color substances that were both brilliant and permanent … the second is the difference of opinion as to the exact color indicated by any name, and the lack of any authority by which an individual opinion can be upheld. … the name Lemon Yellow would seem sufficiently accurate as a descriptive term, yet the color of lemons varies slightly and the memory for exact color sensations, when the original is not at hand, is often faulty. (Maerz and Paul 1)

Readers of James Joyce’s Ulysses may remember the color lemon-yellow, and lemons themselves, as leitmotive that appear at intervals in the novel. First appearing in the Telemachus episode as the “Paris fad” for tea which Buck Mulligan rejects in favor of “Sandycove milk” (Stephen has just recently returned from Paris, and had aquired some of its habits), the color appears in “Proteus,” as Stephen muses about the effects of sunlight on the color of the houses: “Gold light on sea, on sand, on boulders. The sun is there, the slender trees, the lemon houses. Paris rawly waking, crude sunlight on her lemon streets” (Joyce, Ulysses 10, 35). Neither the Sandymount houses nor the Paris cobblestones are painted lemon-yellow, of course, or appear so at all other times of day, but they look this way under the reflection of the early morning light. Stephen, a poet, is more interested in the phenomenology of the visual experience than its lexicon—one which would describe the houses by the name of their paint, or the stones as gray. Lemon-yellow, then, is the site at which the Aristotelian conception of color–that the stones are gray—meets Newtonian color phenomenology—that they appear gray.

Leopold Bloom, too, the hero of Ulysses, imagines the skin of his naked body in the bath, as “lemonyellow,” not because he is jaundiced, or of olive-toned Mediterranean complexion, but because he imagines the light catching his body, “oiled by scented melting soap,” the lemon-scented and lemon-colored soap he’d just bought (Joyce, Ulysses 71). When Bloom later notices the scent of “citronlemon” in his handkerchief, he conflates the citron, an ancestor of the lemon and the French word for lemon, with Israel Citron, a real Dubliner about whom he had been thinking two paragraphs earlier (Gifford, Ulysses Annotated 74, 133). Don Gifford suggests that Bloom “associates the soap with the citron (Ethrog) central in the ritual of the Jewish Feast of Tabernacles (Sukkoth)” (Gifford, Ulysses Annotated 133). In the surreal dream of the Circe episode, this soap appears reified an a sun-god, “diffusing light and perfume.” and speaks in terms of light and reflections: “we’re a capital couple are Bloom and I. He brightens the earth. I polish the sky” (Joyce, Ulysses 340). For Bloom, colors like lemonyellow are a phenomenological crucible where visual experience and olfactory memories are melted together.

But not only are these textual perceptions problematic, but, as Maerz and Paul remind us, the color of lemons themselves varies. In fact, lemons themselves are green before they ripen, and green in certain varieties. In French, a language in which Stephen often daydreams, lemons and limes are citrons and citrons verts, (“green lemons”) most commonly, meaning that lemons can be both yellow and green, in that language’s taxonomy. However, the color lemon, in English and in French, invariably refers to a bright yellow, despite any variation in its actual color. This is a theoretical problem now, but will become a practical problem, in the section below, on modeling color categorization. Everyone knows that lemons are yellow, blood is red, and the sea is blue. But lemons are also green, blood is usually brownish, and the sea may appear purple, brown, or green. So description, then, is both a representation and a social contract.

This fact is one of several properties of color words which complicate the algorithmic design below in the experimental design section, and which bear discussion, as they inform the way we must treat color words, going forward. Three problems in particular deserve attention. The first is that color terminology is perceptually uneven. The second is that colors are indistinguishable from their object-archetypes. And the third is that color descriptions are color exceptions: they describe aberrant color phenomena. Let’s examine each of these in turn.

Perceptual Unevenness: On the Impossibility of a Bluish Yellow

An even more troubling, and more deeply epistemological problem, Ludwig Wittgenstein articulates in his late work Remarks on Color. He asks, quite simply, whether it is possible to imagine a “bluish yellow”:

If you call green an intermediary colour between blue and yellow, then you must also be able to say, for example, what a slightly bluish yellow is, or an only somewhat yellowish blue. And to me these expressions don’t mean anything at all. But mightn’t they mean something to someone else? (Wittgenstein and Anscombe 20e)

Wittgenstein then asks whether a “reddish green” or other color combinations might be difficult to imagine, and why. He posits that the category of green is what prevents him from imagining “bluish yellow,” since, he says, “for me, green is one special way-station on the coloured path from blue to yellow…” (Wittgenstein and Anscombe 22e). This is an important question, with many implications. First, what colors are there which have greater primacy among speakers of English? And more generally: why do linguistic categories—color words and their weights in our language—transform our imaginative processes?I say “our” here with some hesitation, since I suppose an affinity with others who might experience color terminology in the same way, but recognize that a painter, with years of experience mixing colors, might imagine these terms differently, as would, most likely, a speaker of a language very different from English. Still differently would a blind person imagine these colors.

Saija Kerttula calls this the derivational productivity of a color word, and lists a number of ways that a color term may be derived (Kerttula 89). Adding -ish is one, but there is also -ied, as in rubied, -y as in bluey, -ate as in roseate, off- as in off-white, and many others. For the sake of simplicity, however, let’s stick with the most common, -ish, and attempt to test this question of Wittgenstein’s, by examining patterns in literary data. To test this, I constructed a matrix of color expressions from the \(CM_X\) color map,Described in more detail in the \(CM_X\) section below.

where one word ends in -ish. The resulting matrix is shown here in fig. 9.

Not only are there no entries for bluish yellow or reddish green here, but a few other patterns are apparent. First, yellowish green is not mapped to the same color as greenish yellow, indicating that the order of the adjectives dictates precedence. Second, those colors that take -ish adjectives are common colors. However common a color like maroon might be, reddish maroon does not appear in this list, potentially because it’s not considered a basic color with the ability to be mixed. However, some colors which are common in marketing, like beige and teal, but which are less common in paint names, are present here.

Also note that orangeish and orangish, variant spellings of the word, have different average colors here, and orangish is used as a modifier half as much as orange is modified by an -ish. We might say, that orange can -ish, but it is not very -ishable. Greenish and brownish are much more versatile as modifiers than others: they are good -ishers. But green is much more easily -ished than other colors. So does green take a first place in our cognitive pantheon, despite being a secondary color?

Pink has many variations here, despite being simply a shade of red. We don’t see these same patterns with an analog of pink in other hues, like light blue or light green. This leads one to believe that pink’s monolexemic and monosyllabic advantage over analogues like light blue give it more cognitive-categorical weight.

The unevenness of these compound color words reflects similar patterns with visual perception. As early as 1892, the psychologist of color vision, Ewald Hering proposed what he called opponent-process color theory to describe the primacy of certain color categories, and their mutual exclusion (Turner 130). Hering’s theory posits opponent color pairs black-white, green-red, and blue-yellow. The activation of one color cancels its pair. This theory not only helps to explain certain types of color-blindness, but a number of other experimental observations, such as after-images. The competing theory of trichromacy, put forth by Thomas Young and Hermann von Helmholtz in the mid- to late-nineteenth century, posits that color perception is owed to three retinal receptors with different, but overlapping, frequency sensitivities: a red, a green, and a blue. The current understanding of color perception borrows from both of these nineteenth century theories, but acknowledges that color perception is even more complex than these theories would suggest. To some degree, these theories are baked into the equations which the CIE uses to describe perception, as in eq. 3, which I use in my algorithmic design below.

Colors are Intertwined with their Object-Archetypes

Objects and their colors define each other. In a metaphoric sense, objects are colors: we use things to describe our visual experience. Alternatively, our color categories are shaped by the kinds of things we often see which are seemingly illuminated by these colors. The sky is blue is a statement of such obvious fact that the phrase has come to emblematize obvious facts themselves. The sky is blue, leaves are green, roses are red, and violets are blue. Or are they violet?

Put differently: are violets named such because they are violet in color, or is the color word violet the name for the color of the flower? Lexical data from the OED show that violet the color appears at least a hundred years after violet the name of the flower—in 1430 and in 1330, respectively (“Violet, Adj.”). We already know this about mauve. Similarly, we might ask, which came first, the color orange or the orange fruit itself? And there, too, the name of the fruit, taken from the name of its tree, is ultimately descended from Sanskrit, and is older than English itself, but the color sense only appears in the early 16th century (“Orange, Adj.”). So we might deduce that the polysemy between these colors and their associated objects is caused by the phenomenon of naming colors after common objects.

But not all color words are metaphors, of course, and in other cases, the relations between a color and its typical objects is much more complex. A look at some frequently-occurring bigrams shows a kind of magnetism between objects and their appearance. Kerttula calls this a color’s referentiality, or application (Kerttula 87). Some color words we expect to see describing certain objects, and certain objects recall their common color descriptors.

To measure this, I designed the experiment below, in order to find out what objects are most often described as blue, red, and so on. I wanted to quantify the gravitational pull of colors with their associated objects. I use data from the Google Books Ngram Viewer project, hereafter \(C_{NG}\) (Google Books Ngram Viewer Exports). Google Books provides n-gram (sequences of words of length \(n\)) data for the books in its vast collection, and the most recent version, 20200217, provides n-grams tagged according to their parts of speech. I use the data subset English Fiction, which, although not strictly relegated to the time and place this I am studying here, is still useful to determine broad patterns.I describe this corpus in greater detail in the appendix.

There, I query for patterns ADJ NOUN, where ADJ is a color word, tagged as an adjective, and NOUN is any noun which follows. The list of color words I derive from Berlin and Kay, but augment with several auxiliary color words, for comparison (Berlin and Kay).

Fig. 10 shows word cloud visualizations for each color word and its most commonly collocated nouns.Word clouds, or tag clouds, are a relatively recently popularized technique of textual data visualization, which depicts the frequency of words through typeface sizes. See Viégas and Wattenberg for a history of the visualization that traces it to Soviet Constructivism.

Surprisingly, the most frequent collocations are not always the most cliché: green grass and green leaves are of course present, but are not as frequent as green eyes. Blue sky is present, and red blood, but are subordinate to hair and eyes.

white

black

red

green

yellow

blue

brown

purple

violet

pink

orange

gray

greenish

reddish

bluish

Some overall trends are apparent in these words, which may be illuminated by categorizing them. Using WordNet, the relational lexical database, I am able to identify hypernyms for most of these words, and group the words according to these hypernyms (Miller). The hypernym treemap for red, for instance, shown in fig. 11, shows that a good proportion of the words are body parts (they have the hypernym synset body_part.n.01), or “coverings” such as hair (covering.n.01). A crucial similarity between these bodily descriptors, from red hair, the most frequent collocation, to red eyes, red lips, and red face, is that they describe exceptions, or aberrations, from their usual states. Red hair (actually orange, as I will show) is among the least common natural hair colors. Red eyes describe diseased or depigmentized eyes, of humans or other animals. And red lips are lips that are unusually red: whether blood-filled through vigor, or excitement, or through the use of cosmetics.