Color and Color Perception II

Color in Art and its Literature

The Formalist Tradition

One of the longest-running debates about color has concerned its cognitive status; ever since Antiquity, there has been a fairly clear-cut philosophical division between those, like Berkeley or Goethe, who considered that our knowledge of the world was conditioned by our understanding of its colored surfaces, and those, like the ancient skeptics or Locke, who regarded color as an accidental attribute of the visual world, and visual phenomena themselves as an unreliable index of a substance. Cézanne’s career as a painter might well be characterized as a sustained meditation on this theme. There is now some reason to think that there may be a biological basis for the belief that tonal variations in a scene supply the viewer with most of the information needed to interpret it. Monochromatic engraving and photography are the most obvious manifestations of this belief in Western art; but it is a belief that would also help us to understand the persistence of a light and shade (value) based color-systems in the West, from Greek Antiquity until the nineteenth century, as well as the recurrent debate on the respective places of disegno and colore in painting, a debate that took a particularly philosophical turn in the seventeenth century, when, especially among Italian artists and theorists, the cognitive independence of line and ‘form’ was increasingly claimed.

As it happens, the only ‘school’ of color-analysis in the history of art owes its development, not simply to the theoretical framework proposed in the nineteenth century by the Swiss critic Heinrich Wölfflin, who focused on the formal characteristics of visual style, including color, but also, and more importantly, to the stimulus of the more recent philosophical tradition of phenomenology, represented in Germany chiefly by the philosopher Edmund Husserl. Lorenz Dittmann’s wide-ranging study, Farbgestaltung und Farbtheorie in der abendländischen Malerei (Color structure and Color theory in Western Painting), 1987, is only the most important summation of a tradition of Koloritgeschichte (history of color in painting) that goes back in Germany to around the time of the First World War and has engaged a considerable number of distinguished art historians, including, most recently, Theodor Hetzer, Hans Sedlmayr, Kurt Badt Wolfgang Schöne and Ernst Strauss.

Husserl’s pupil, Hedwig Conrad-Martius, took the study of the phenomenology of colors out of the psychological laboratory and into the studio and the gallery; away from a concentration on nature and into paintings, where nature was exposed in all its chromatic wholeness.

Conrad-Martius’s color-theory [wrote Dittmann] shows us again [i.e. after Goethe] that only a developed nature-philosophy, a comprehensive ontology, will be fruitful for the perceptive, thoughtful engagement with works of art. An isolated ‘aesthetic’ will hardly serve, and only occasionally the individual sciences, such as experimental psychology, which are tied down to the empirical.

Within the framework of a phenomenological study of color in art, the role of light and shade (values) and the role of chromatic elements (hues) have been particularly difficult to distinguish, and it is not surprising that the only classic survey of the field remains Wolfgang Schöne’s Über das Licht in der Malerei (On Light in Painting, 1954), which is still being reprinted after four decades. The hallmarks of this school of analysis are an immediate confrontation with the object, and a systematic and sophisticated technique’ and terminology for describing the effects of that confrontation. In his large-scale study, Dittmann has been somewhat dismissive of the work of his only rival in the field, Maria Rzepifiska, whose History of Color in European Painting, he claims, neglects the comprehensive study of individual works. But as his own work shows, confrontation is a method that is fraught with pitfalls. Many of his sensitive analyses are masterly for example, the paragraph on Millet’s The Gleaners, which gives a very precise sense of the way in which the almost indefinable, shifting nuances of the painter’s palette nonetheless contribute to the creation of a stable, monumental structure:

The Gleaners. . . is dominated by a muted brightness with a brownish and grey-violet undertone. The sky appears tinged with a tender violet, as it were a mixture of the two most evident hues in the picture: the very dull grey-blue and copper-red tones in the headscarves of the bending peasant-women. In the white sleeve of the central figure, the light gathers with a ‘filtered’ effect. The unusually restrained colors (which seem to contradict the monumental forms) follow a closely-stepped sequence: reddish tones in the central figure, based around copper-reddish, brownish and bright carmine; delicate nuances of colorful greys in the standing figure to the right: silvery bright blue-grey, dove-grey, blue and turquoise greys. The color-thresholds are kept so low that induction effects are made much easier, which allows the indefinite color-tones to appear as ‘resonances’. Thus the barely definable, shimmering brownish tone of the field in the middle distance takes a tender pink-violet tone against the grey-scale of the figure at the back, which echoed again in the slightly darkened foreground.

But a lengthy book made up of such plums, particularly one for which the publisher has (justifiably) chosen the austerity of an unillustrated text, would indeed indigestible, and there are fortunately not many set pieces like this. In any event, Ditttnann soon gets into trouble with his principle of personal encounter, he has simply not been able to examine in the original all the artifacts he wants to discuss. His chapter on medieval book illumination — that most inaccessible of art forms, rarely available for inspection, and usually displayed one opening at a time depends entirely on descriptions by Heinz Roosen-Runge: and indeed, Dietmann’s text, in general, owes much to quotations from other scholars such as Theodor Hetzer and Kurt Badt, and of all, Ernst Strauss, whose unpublished notes well as published works (which Dittmann edited) have provided him with a good deal of material. But the visual analysis of color can in principle never be at second hand, for different eyes will, as like as not, see quite different things.

This type of detailed visual analysis works well enough for gallery paintings such as Millet’s; far more disturbing than the occasional reliance on informed hearsay is Dittmann’s almost complete disregard of the context of seeing. The discussions, for example, of Taddeo Gaddi’s frescoes in the Baroncelli Chapel of S. Croce In Florence, or Ghirlandaio’s in the choir at Sta Maria Novella,  do not so much as mention the stained glass in the windows which gives a pronounced coloristic atmosphere to the architectural space; and this is the more surprising in that Schöne had devoted a good deal of attention to the problematic effects of environmental light (Standortslicht), particularly in the context of the modifying effects of stained glass on the frescoes in the Upper Church of S. Francesco at Assisi.

In dealing with the painting of the nineteenth and twentieth centuries, Dittmann gives less and less space to his own visual analyses and more and more to the statements of the painters themselves, even to the extent of reprinting Delaunay’s short essay on light in both its French original and the German translation by Paul Klee. A belief in the primary importance of artists views of their own color practice is also a notable feature of the approaches of Strauss and Badt, whose studies of Delacroix and van Gogh depend heavily upon those painters’ abundant writings. But we are increasingly aware that painters are not necessarily privileged spectators of their own works; and when they turn to words they may in fact be rather less able than other categories of writer to articulate their thoughts about the notoriously opaque world of visual sensation. One cannot but be struck, for example, by the poverty of idea and expression in, say, Mondrian’s writings between 1917 and 1944, or Matisse’s between 1908 and 1947, compared to the richness and variety of the work to which these writings ostensibly relate. In the case of Matisse we are dealing with a far more sophisticated thinker than Mondrian, but the simplifications that arise from an essentially propagandistic intent are no less evident’ Artists’ statements are not transparent; they must be unpacked like any others. For example, it would have been helpful to have had Dittmann’s commentary on the manifest differences of tone and emphasis of Delaunay’s essay La Lumiere, and Klee’s version of it, in which Delaunay’s loosely structured meditations on the primacy of sight and the spatial effects of light which created what he called ‘rhythmic simultaneity’ were given a far more coherent structure and a far greater on the complementarity of polar energies.

Though my heading above characterized the German school of Koloritgeschichte rather crudely as ‘formalist’, it is clearly not formalist in any rigorous sense. Since it grew out of Conrad-Martius’s theory that color serves to identify the very essence of being, it could hardly have rested content with the mere analysis of external characteristics. It is true that one of the few classical aecheologists to have been affected by this approach, Elena Walter Karydi, has undertaken the improbable task of draining the symbolism even from archaic Greek color. But the search for literary ‘meaning’ in color has been pursued by followers of this tendency, not only where we should most expect it (for example in Uwe Max Rüth’s dissertation, Color in Byzantine Wall-painting of the Late Paleologian Period, 1341-1458), but even in Gisela Hopp’s monograph on Manet — a painter whose style has until very recently been interpreted as a ‘realist’ ancestor of Impressionism, and hence largely free of literary or symbolic content. Hopp’s treatment of expressive color in Manet is particularly interesting because in her analysis of a number of the major canvases she makes much of the painter’s use of emerald green, a pigment that in German has quite deservedly been named ‘poisonous green’ (Giftgrün). In The Balcony, Hopp saw this green as overwhelming and oppressive,  and in her discussion of the late Bar at the Folies-Bergére she was even tempted to identify the characteristically bulbous bottle on the bar to the right as holding green absinthe, and, by contrasting with the ‘heated orange’ next to it, helping to establish a mood of tension and irritation in the picture. But, as Françoise Cachin noted in her account of this painting for the Manet exhibition of 1983, the green bottle contains, not absinthe, but the far cozier creme de menthe, which is still marketed in this format. The creme de menthe sits very well with the equally identifiable bottle of English beer. Perhaps Hopp’s interpretation of the greens in Manet’s paintings was largely conditioned by her use of this particular German term, and thus raises the question, to which I shall return later in the chapter, of how far symbolic interpretation may simply verbalize a visual attribute.

Central to the problem of formalism in this style of color-analysis is its relationship to a notion of history. Dittmann’s meticulous and highly selective method resists historical generalization; and Schöne has stated quite categorically that the starting point of any investigation must be the impressions made on the modern investigator him- or herself.  It is not at all surprising that there is a certain lack of historical dynamics in this sort of writing. Dittmann, to be sure, makes historical judgments from time to time, for example, that the seventeenth century saw the fullest development of chiaroscuro, or, less plausibly, that color in the twentieth century gained quite new independence in art. But these judgments are quite ancillary to the detailed characterization of a selection of ‘key works’. Sometimes Dittmann is struck by what seems to him to be the earliest significant use of a particular hue. Brown a particularly interesting case in point. As a non-spectral color, brown has been especially resistant to theory, and philosophers and experimental psychologists have generally argued that it is simply a darkened variety of spectral yellow. But, although it may be perceived to be unmixed, brown also has a very wide range of affinities with the long-wave spectral colors yellow, orange and red. Traditionally, and some European cultures until remarkably recently, it has had, like blue in earlier periods, the general connotation of ‘dark’. Because of its importance in painting, brown has particularly attracted the attention of the German school of color-analysis, beginning at least with Conrad-Martius. Dittmann traces the ‘diccovery’ of brown as a unifying pictorial device to the late Quattrocento in the work of the Pollaiuoli and Signorelli, but other scholars have dated its coming of age to the early work of Velazquez and Ribera. The identification of this rather late emergence of brown is given a certain force by the undoubted conceptual link between brown and darkness in the seventeenth century (and in French, for example, brun still means dark), but it is also supported by the evidence of Iberian treatises on painting in this period, which list an exceptionally large number of earth-browns as habitually in use. It is contextual material of this kind that is needed to turn visual analysis into history.

The Substance of Color

Koloritgeschichte is notable for a certain reluctance to consider the material condition of the works of painting it to analyses. Yet perhaps the most important developments in the study of painterly color in recent years have come from conservationists, who have been making the results of their campaigns increasingly available to the general public as well as to historians of art. Technical discussions have become commonplace in exhibition catalogues dealing with all periods of art, and there have been several popular exhibitions on restoration itself.⁴⁸ It is particularly remarkable that the specialist literature of conservation, such as Studies in Conservation or Maltechnik, has now been widely supplemented by periodicals that are clearly aimed at a general readership. Catalogues of single artists as well as catalogues of particular collections are now likely to be provided with far more technical information than hitherto.

Not that conservation is likely to give formalist critics much joy: the enormous help that it can give in matters of connoisseurship is hardly matched by its aid to aesthetic presentation (see, for example, Leonardo’s newly stripped-down Last Supper in Milan); and conservation methods are, of course, a very controversial area among historians of art as well as among conservators themselves. In recent years the cleaning of some Titians at the National Gallery in London, the restoration of the glass of Chartres West, and, most of all, the cleaning of the Sistine ceilings have given rise to much excited debate, which is not, since it is primarily a question of aesthetics,  ever likely to reach any settled conclusions. What restoration reports do offer the historian of color is more reliable information than that hitherto available about the methods and materials of painting in many historical periods — methods and materials that have often been part of an ideology or mystique of technique specific to those periods and to particular places. There have been a number of recent exhibitions devoted to techniques and materials, but rather less attention has been given to tools. The pioneering work of E. Schimid and J.W. Lane and K. Steinitz fifty years ago on that most important practical conceptual as well as practical tool of the artist, the palette, has only recently been developed.

Color symbolism itself has sometimes been thought to depend on the qualities Michael Baxandall pointed to the way in which certain Florentine contracts of the fifteenth century prescribed specific qualities of ultramarine for the most important areas of the picture, such as the Virgin’s robe because it was the of all pigments; and this is an attitude also found in seventeenth-century Spain. Yet, as both contracts and the technical analysis of surviving works abundantly show, other blue pigments were used as frequently these vital places, and the most important Italian recipe book of the period described synthetic blues that were claimed to be indistinguishable from the best ultramarine. Contracts often specified other particularly expensive pigments, as well as gold, and the use of these specified colors was prescribed by many Italian guild regulations: rather than demonstrating a ‘materialist’ attitude to color symbolism in the spectator, they show a concern for the color-stability of the product, which, it was assumed could only be guaranteed by the use of the ‘best’ materials.

One of’ the important conclusions to be drawn from much recent research in conservation is that artists’ practice at all periods was often far more complicated than the handful of surviving technical texts would suggest; and with the exception of Roosen-Runge’s study of the Mappae Clavicula and ‘Heraclius’ texts in relation to English Romaneque manuscript illumination, David Winfield on Byzantine mural-techniques, and Mansfield Kirby-Talley’s account of the theory and practice of the eighteenth-century English portrait-painter Thomas Bardwell, there has, it seems, been little attempt to test the texts against the practice. Nor has the corpus of written texts expanded much in recent years, although there have been important new editions of the standard ones.

A systematic survey of scientific sources, particularly medical literature, would certainly extend the range of technical sources for the arts; there is, for example, some particularly rich material on dyeing and painting including what appears to be the earliest textual reference to oil painting in a recently published treatise on the elements by the southern-Italian physician Urso of Salerno, dating from the late twelfth century. The vastly expanding technical literature for artists in the nineteenth century has still to be and evaluated, although Anthea Callen has used some of it in her important study technique. Virtually no work at all has appeared so far on the technical interests of twentieth-century painters, although the commercial development of new artists materials has been greater our time than any earlier period, and they have as usual formed an important part of the prevailing aesthetic ideology.

Theories and Assumptions

It has been quite a common practice among writers on color in art to preface their analyses with an account of color-phenomena in general, an account for the most part bawd on the literature of experimental psychology of the past century or so. It is quite unrealistic to suppose, however, that the psychology of color perception has reached firm ground, and its relationship to the practice of painting must thus remain highly problematic. The art historian must, I think, be more concerned with the local context of color-ideas as they relate to the artist under consideration than with any global theoretical framework; and in many cases, these ideas will be assumptions rather than anything that could be plausibly be presented as a theory. The treatment of color-theories has usually been the weakest element in the discussion of what may lie behind the choice and handling of color in a given artifact; and this has been because historians of art have found it hard to shake off that ‘progressivist’ approach to their subject which historians of science have long since discarded. They have tended to expect more coherence in the handling of theory by painters than the evidence would warrant, and to see in the color theories of the remoter past a unity and simplicity that in most cases have barely been achieved even today, as well as a tighter fit with practice than it is reasonable to expect. This does not make color-theory any the less important.
Stated in its broadest terms, the theory of color in the Western tradition, from Antiquity to the present, can be divided into two phases. Until the seventeenth century the main emphasis was on the objective status of color in the world, what its nature was, and how it could be organized into a coherent system of relationships. From the time of Newton, on the other hand, the emphasis has been increasingly subjective, concerned more with the understanding of color as generated and articulated by the mechanisms of vision and perception. At the same time, the relationship of science to color has shifted from an earlier dependence of scientists on artists, who in their capacity as technologists of color supplied science with the necessary technical and experimental data, to an increasing dependence, about the end of the eighteenth century, of artists on scientists, whose growing professionalism and prestige allowed them to offer more, and more that was beyond the reach of art. Even the early treatises for artists, such as Theophilus’s De Diversis Artibus or the slightly earlier anonymous De Clarea, can now be seen, not merely as random collections of recipes, but as incorporating often quite sophisticated statements of theory. Conversely, it is very hard to find artists capable of absorbing the color science of any period after the early nineteenth century. On the other hand, attempts to reconstruct a philosophical context for ancient color-practice — attempts that go back at least to the eighteenth century but are still an active preoccupation of Classical scholars have not been able to overcome the brevity and unreliability of the written sources and the ambiguities of the surviving monuments. As Leon Battista Alberti noted in De Pictura, and in support of his own literary efforts, several ancient artists had written on painting, but none of their writings has survived, and we are still dependent largely upon the architect Vitruvius and on Pliny for our interpretation of the styles of the earlier Greek examples of painting that are coming increasingly to light. The key text has always been Pliny’s account (Natural History, XXXV) of the four-color palette of Apelles and some of his contemporaries, which has been related to an archaic Greek doctrine of the ‘basic’ colors of the four elements. While several scholars have continued to use Pliny’s text as a guide to color-principles in the fifth and fourth centuries BC, others have more plausibly placed it, with the related judgments of Vitruvius and the orator Cicero, in the context of a specifically polemic of against extravagance in decoration.

Alberti to Dürer

Alberti’s Dc Pictura, which includes a number of important remarks on color, was an entirely new kind of theoretical text in which practicalities played a very minor role, although the author was also a painter. It has suffered from being seen, in its emphasis on ‘variety’ and on the tonal scale, as embodying a very medieval attitude toward color, and as depending more or less exclusively on Aristotelian tradition. Rather little has been published so far on specifically fifteenth-century developments in optics, but Alberti’s interest in the effect of light and shadow on colors can be paralleled in some contemporary Central European, if not Italian discussions, and anticipated the far more extensive investigations by Leonardo at the close of the century. Alberti’s remarks on the harmonious assortment of colors in painting also reflect a preoccupation in Florence in the early fifteenth century. Lorenzo Ghiberti’s Commentaries, and particularly his Third Commentary, may have been stimulated by Alberti’s work, although they were never shaped into a coherent treatise. But where Alberti was content to leave to ‘the philosophers’ the detailed discussion of the nature and effects of colors, Ghiberti drew heavily on these same (mainly medieval) philosophers, so that his Third Commentary is as it stands little more than an edited selection of passages from earlier authors.  But, this does not detract from its relevance to Ghiberti; practice, especially as a jeweler and a stained-glass designer.

Leonardo, too, looked very widely at medieval writers on optics, but he found their opinions difficult to reconcile with his own experience and the results of his experimentation. The work of Corrado Maltese in the 1980s has sought to weld some of Leonardo’s scattered remarks on the mixture of colored lights into a more or less coherent prefiguring of the modern theory of additive and subtractive mixture; but although Maltese recognizes the many gaps in the painter’s experimental procedure, he has still tried to fill too many of them with his own engaging speculations.

His argument that in the course of his work Leonardo was able to reduce the traditional four-color scheme of ‘simple’ colors to the modern three, flies in the face not only of Pedretti’s revised dating of the Codice Atlantico, where much of this work appears but also of Leonardo’s frequently changing attitude to what constituted a ‘simple’ color both green and blue, for example, appear as compounded colors In various notes. Least convincing of all is Maltese’s attempt to link Leonardo’s perception of the formation of colors through semi-opaque media with the glazing methods used in the underpainting of the Ufizi Adoration and the Vatican St. Jerome.

What seems increasingly clear is that Leonardo’s inability to elaborate coherent theory of color, and his traditional distrust of the capacity of color to reveal the truth, fuelled his inclination to regard light and shade as the primary visual phenomena, and simulated his development of techniques in drawing and painting to exemplify this truth. Recent commentators have underlined Leonardo’s view of the dynamic power of darkness, superior even to that of light, and his creation of a new and fruitful concept of chiaroscuro Leonardo was even suspicious of bellezza beauty because for him it implied lightness. His supremely subtle interpretation of chiaroscuro in art, and particularly his technique of sfumato, was to involve unprecedented experimentation with media, including the development of soft pastels, and the extensive use of those most delicate and responsive of all painting tools, his fingers.

Much has been made of the little that I hirer wrote on color effectively only a short note on drapery-painting. which he advocated modeling without ‘shot’ effects, advice that, as Dittmann has pointed out. Dürer was not always inclined to follow himself. More promising, perhaps, is the linking of Grünewald’s unearthly color with his experience of the theory and practice of metallurgy, although there are no indications so far that Grünewald ever turned to color-theory as such.

Although the sixteenth century unusually productive in color-theory relating to the arts, little of it was by or addressed to painters, and it to have borne only tangentially on their practice.

Science into Art

Only around 1600 did the theory of color seem to offer something new and exciting to artists, and the widespread movement to integrate the art and the science of color, which began essentially at the court of Rudolph Il in Prague, was to last for nearly two centuries. In the era of the Kunst- and Wunderkammer, color and colors, like painting and engraving, were among the wonders of art to be set beside the wonders of nature. In Rudolph’s entourage several artists and scholars –the painter Arcimboldo, the mathematician Kepler, the physicians de Boodt and Scarmiglioni -were interested in color, and especially in its relationship to music. During the seventeenth century many artists became involved in color-theory, and many theorists of color looked to painting for enlightenment. It was the period when Leonardo’s writings were first evaluated and published, and when artists in both northern and southern Europe turned their hands to writing. There are now studies of the theoretical interests of Rubens, Poussin, and Pietro Testa, as well as of the minor painter but influential theorist Matteo Zaccolini. In the burgeoning French Academy of the 1660s color and its relationship to design became a standard topic of formal, as well as informal debate, generating an important and influential literature, especially by Félibién and De Piles. Paradoxically, since this was also the period that gave the greatest value to darkness, both in theory and painterly practice, light and color found for the first time a unified theory in the work of Descartes and, especially, Newton, who showed that color was indeed illusory, and that light was its only begetter. Yet artists were at first both willing and able to draw on Newton’s ideas, especially his conjectures about harmony, and his circular arrangement of colors which eventually gave them a clue to the nature of ‘complementary’ contrast. Contrasts are, of course, subjective effects, and it was one of the greatest achievements of Newton to have shown that all color is intrinsically subjective.

Science — ‘the taste of all minds’

After Newton, the aspects of color-theory most interesting to artists have been, in addition to theories of harmony, the devising of color-systems, and the exploration of how colors relate to the mechanisms of perception, and affect the feelings of the spectator. Many of these concerns had long since developed in artists, studios themselves, but they were now investigated and codified systematically. Long before the Czech physiologist J. E. Purkinje announced the principle of chromatic shift in subdued lighting that now bears his name, it had been part of studio lore that paintings could best be examined in the conditions of lighting in which they were made.

The most comprehensive contribution to the study of color, which laid great emphasis on these subjective phenomena, was Goethe’s Farbenlehre (Theory of Colors) of 1810. It has a claim to being the most important single text on color for artists, and, indeed, for historians of art, since one part is devoted to ‘materials for a history of color’, including what appears to be the first historical outline of color in painting, contributed by the painter Heinrich Meyer. Recent studies of the Farbenlehre have tended to revive the old theme of Goethe’s outspoken opposition to Newton’s theory that color is a function of light alone, this campaign in favor of Goethe has ceased to be the preserve of spiritual movements and has joined the mainstream of the history of science. Newton is, of course, no longer the rationalist idol he was in Goethe’s day: we have long had Ronald Gray’s Goethe the Alchemist, and we now have Newton the alchemist, although Newton’s practice of alchemy has hardly been brought to bear on the history of his optics, as Goethe’s has. What is more important for us is the puzzle of why a theory of color so patently directed at artists, and deriving partly from Goethe’s theoretical and practical experience of art, should have made so little impression on artists for nearly a century. The original and extensive theory of Philipp Otto Runge, the one painter who was close to Goethe during the final stages of his color-work, has in detail very little to do with Goethe’s theory. Unlike Goethe, Runge was unable to develop an integrated theory: his published Farben-Kugel (Color Sphere) of 1810 was in the tradition of European colorimetry in the seventeenth and eighteenth centuries — the direct descendant, indeed, of the 1611 color-sphere
of the Swedish mathematician Sigfrid Forsius, while his, unpublished thoughts were in the metaphysical tradition of the late-Renaissance speculators G. P Lomazzo or Athanasius Kircher. Neither Runge’s practical experiments with transparency. which may be related to his delicate watercolor technique, nor his ideas on color-meaning, which he sought to exemplify in the unfinished cycle of the Times of Day, bore fruit in the rather austere diagrammatic format of the Kugel, although Matile has shown how Runge brought his published ideas, of harmony to bear on the small version of Morning in the Han)burg Kunsthalle.

Whereas in the seventeenth century the scientific theory of color drew largely on the experience of painting, not least in the search for a set of ‘primary’ colors, by about 1800 the balance had shifted, and the very extensive development of scientific color-theory since Newton was now directed to artists through many popularizations, sometimes at the request of the artists themselves. This does not mean that artists were invariably willing, or even able, to use the color-information supplied to them in this way; and the more circumspect studies of the relationship of color-theory to painting in the nineteenth century have shown that theory and practice very rarely went hand in hand. But then we should not have expected that theory, any more than ‘nature’, would have been ready for direct and complete transposition into art. The extraordinary vitality and tension of much nineteenth-century coloristic painting derives precisely from the struggle with the intractable ideas and sensations of color. The Newtonian solution to the problem of an antithesis between ‘apparent’ and ‘material’ colors had thrown the scientific emphasis entirely on to the study of light, and decisively separated the procedures of the laboratory from those of the painter’s studio. Runge, who experimented in both traditions, remained hopelessly if fruitfully, confused about the relationship of theory to practice. Most painterly theory in this period was more or less anti-Newtonian, and it is not surprising that Turner, for example, felt himself drawn even at an advanced age to study the theory of Newton’s leading opponent, Goethe. But Turner’s theory of color mingled traditional and modern elements in a thoroughly idiosyncratic way, and it remains an open question how far he understood the main issues at stake.

Delacroix’s thoughts on color have also generally been linked with the publications of a single theorist, the chemist M.—E. Chevreul. But, for example, the well-known color-triangle with a note on primaries and secondaries in Chantilly derives from a less abstruse source, J. F. L. Mérimée’s De la Peinture a l’huile of 1830, and it was not until about 1850, when Delacroix was deeply involved with the technical problems of large-scale ceiling-painting, that he seems to have turned to Chevreul for advice, acquiring a set of notes from a lecture-series of 1848, and proposing to visit the chemist in person. It was at this time, too, that Delacroix came to know Charles Blanc, whose Chevreulian interpretation of the painter’s color-handling served to assure the younger generation of the 1880s that Delacroix was indeed a ‘scientific’ colorist.

Blanc was perhaps the most important of the mid-nineteenth-century French writers on color because he was read so avidly by Seurat, Gauguin and van Gogh among others. An admirer and follower of Ingres, he regarded coloring, para doxically, as an inferior part of painting; and it was from pupil of Ingres, Jules Claude Ziegler, that he took his color-diagram and, probably, his first knowledge of Chevreul. Like Ziegler, Blanc moved easily between the line and the applied arts- he also wrote a Grammaire des arts decoratifs- and he saw no contradiction in applying the same color-theory to both. With far-reaching consequence, he also praised Oriental cultures, especially the Chinese, as expert in color and models for European color-usage. It was on to an Oriental -albeit a Turk- foisted his amusing pastiche, the brief essay on color harmony that he circulated among some friends in Paris in 1886.

Seurat’s reputation as a theorist has suffered somewhat recent years, and it is certainly not easy to understand why he remained so loyal to Chevreul, when the literature of color for artists in the 1870s and 1880s had introduced the far more sophisticated notions of Hermann von Helmholtz. The explanation may lie in Seurat’s belief in Blanc’s view of Delacroix as, a Chevreulian painter; for it seems that the color-circle that Seurat drew on a sheet of sketches for La Parade is a reminiscence of the circle that Delacroix sketched in a notebook of about 1840, and that had been published by Auguste Laugel in 1869. Laugel’s commentary is interesting for he introduces the new research of Helmholtz into the colors of light, with its scheme of complementaries red-blue/green, orange. cyan, yellow-indigo, yellow/green-violet, but he argues that Delacroix ‘crude diagram’ of Chevreulian complementaries is far more practical for artists. Scurat clearly agreed. The context of Seurat’s scientism has still be fully examined, but it seems likely that in future less emphasis will be placed on the physics of Helmholtz and more on the psycho-physics of Charles Henry.

If Seurat as a color-theorist has been the victim of revisionism, knowledge of van Gogh’s approach to color has remained essentially where Kurt Badt left it in his 1961 study, which gathered a large number of references to color from the painter’s extensive correspondence and related them only very loosely to his work. Vincent’s own writing has continued to be the almost exclusive source of documentation, and although we know a good deal about his reading of the theoretical literature of the period, very little has been done to evaluate his use of it. Nor has the crucial friendship with Gauguin in 1887 and 1888 been looked at closely from the point of view of their rival conceptions of color. Gauguin’s sympathy with Vincent’s notions, shown in the very flatly painted and strongly colored Vision after the Sermon in Edinburgh, and in the lesson he gave to Paul Sérusier in 1888, gave way increasingly to dislike for what he considered to be van Gogh’s very crude color-aesthetic. Although Gauguin never showed much interest in color theory as such, the color-system later published by Sérusier, with its emphasis on warm browns and cool greys and its avoidance of complementarity, may substantially represent Gauguin’s views.

Cézanne’s late work is perhaps the highest exemplification of a nineteenth-century theory of color-perception as a sequence of naively apprehended flat patches, made popular in France by the publications of Helmholtz and his followers. Shiff has drawn attention to this strand of but he has not explored the consequences of these ideas for Cézanne’s style; and the debates continue about
whether Cézanne may be considered to have had a ‘theory’, and the relationship of theory to his painterly practice. Here is one area where formal analysis still has a major role to play.

Twentieth-century Theory

The historiography of color in the art of the recent past has faced two intractable problems. The first is that the categories of color-analysis — the terminology introduced in modern color-systems, and the concepts of the psycho-physiological effects of colors — are the very same ones that have been developed over the past century or so; and they have thus tended to be taken for granted and exempted from historical analysis. The second problem has been the hermetic character of modernist criticism, and, together with this hermeticism, the extensive self-analysis of artists themselves, which this criticism has often seen as sufficient. Criticism, that is to say advocacy, has naturally taken precedence over the more analytical procedures of history. Thus, although the more or less collected writings of some of the major movements, such as Russian Constructivism and the Bauhaus, and some of the major figures, like Matisse and Mondrian, and minor ones, like Marc and van Doesburg, Hans Hoffimann and Winifred Nicholson, are now readily available, there has been remarkably little secondary discussion of the color-ideas of twentieth-century artists.

Several general treatments of individual artists, however, such as Hoelzel, Itten, Matisse and van Doesburg, and of groups such as the Orphists, Russian Constructivists and De Stijl, have included important considerations of their theoretical interests in color. There has also been a handful of short essays on Orphism, on Russian Constructivism, on Marc, Klee, Picasso and Rothko, that have focused on color, and a few monographic studies with the same emphasis.

Several exhibitions in recent years have dealt with color-theory in the twentieth century, or have given a large place to it in the context of some other concern.  What these studies have usually lacked has been some sense of the ways in which the discussions and usages of artists have related to the more general concerns of color-theory in their time. I have made a limited attempt to point to the psychological context of Kandinsky’s, Delaunay’s and Mondrian’s idea, and to the debates on the structure of color-space that form such an important part of early twentieth-century color-science. The wide range of attitudes toward color as dynamic and affective that Kandinsky deployed, for example in his On the Spiritual in Art (1911-12), including a color-system that owes as much to the late-nineteenth-century Viennese psychologist Ewald Hering as it does to Goethe, can be paralleled closely in the responses revealed in a long series of interviews and experiments, chiefly with artists and professional people, conducted by the psychologist G. J. von Allesch in Germany in the decade before the First World War.At the Bauhaus in the 1920s Kandinsky was probably the teacher  most inclined to draw, as we know from his lecture notes, on the recent literature of experimental psychology, notably Neue Psychologische Studien (1926).

The Bauhaus represents a particularly rich field of color-study, where the traditional concentration on the ideas of the most famous of the individual teachers has given quite a false impression of what was actually taught about color there. It never seems to have been a central issue. Itten is assumed to have taught the Course (Vorkurs), compulsory for all students, from the outset in April 1919, but the first prospectus makes no reference to it, and it does not appear in the deliberations of the Masters’ Meetings until October 1920. At Weimar, after Itten’s departure 1923 color was taught in the Vorlehre by Kandinsky for a mere hour a week pared to the eight hours of form-study under Moholy-Nagy plus an hour of the same with Klee, two hours of drawing with Klee, and two of analytical drawing with Kandinsky. Klee’s color-lectures of 1922-3 (excerpted by Spiller in his edition of the Notebooks and now available in facsimile) must have been given to more advanced students in only some of the workshops. After the move to Dessau, under Moholy-Nagy and Albers color appears to have been dropped from the Vorkurs entirely. Albers, however, carne to put color at the centre of his interests, and after his move to the United States he taught the color-course that gave birth to his great Interaction of Color of 1963. In this beautiful and influential book Albers relegated ‘theory’ to the final stages of practice; and it is certainly questionable how far he had a coherent conception of color-theory at all.

Stare for a moment at the red disc, and then, with eye unfocused, at the white disc. Most people will see after-image they would call “blue-green”.Tey since about 1800 red complement has usually been described simply as “green” partly because in the system of the three primaries red, blue and yellow, the complement of each color was deemed to be equal mixture of the other two. The optical evidence is secondary.

Color as Content

In a review of the Titian exhibition in Venice of 1935, and of Hetzer’s book on Titian’s color, which coincided with it, Oscar Wulff accused Hetzer of setting up far too abstract a model of that painter’s color-concerns, and of neglecting color’s Darstellungswert, or representational function. Since Hans Jantzen’s pioneering essay ‘On the Principles of Color-composition in Painting’ of 1913, which introduced the concepts Eigenwert (autonomous function) and Darstellungswert of colors, German scholars have sought to understand the role of color in painting as moving essentially between these two poles. Hetzer himself argued that in the 1530s Titian solved a coloristic problem that had plagued painters since the early fifteenth century: that of striking a balance between the function of color to articulate space (Raumwert) and its surface function (Flächenwert), between its natureas phenomenon (Erscheinung) and as material pigment, between color as beauty and color as meaning. But what these scholars understood by representational or meaningful color was essentially its capacity to imitate the object; that it had any intrinsic capacity to convey meaning they left entirely out of consideration. Wulff suggested, for example, that Titian’s great command of black may have derived from his experience as a portrait-painter rendering the black dress of his many male sitters; but he did not inquire why black was such a high-fashion color in the mid-sixteenth century.

In figure-painting, of course, the deployment of colored drapery has always been a major vehicle for the free exercise of aesthetic choice; and Cennini for example, in a little-noticed passage of the Libro dell’arte, argued that the color-designs of leaves or animals used in block-printed fabrics were entirely a matter of fantasia, provided they created an appropriate contrast. Here, of all places, we might expect color to be unencumbered by any but formal considerations. And yet textiles are perhaps the colored artefacts most expressive of social values, and, through these values, of ideas.

Color- Change: Shot Fabric and Modelling

Historians of textiles and costume have not yet given much attention to question of color, and historians of art have so far used costume almost exclusively as an aid to dating. There has indeed been a tendency to treat the handling of color composition in painted draperies as if it were entirely an aesthetic matter. The art-historical treatment of ‘shot’ materials (cangianti or changeantes), is particularly instructive. Since Siebenhühner’s study of 1935, several historians of Italian Renaissance art, notably John Shearnman, have discussed the technique of color modelling by hue rather than value-shifts, resulting in effects that seems close to those of silk woven so the weft of one color is dominant when seen from one direction, and the warp of a contrasting color is dominant viewed from another. Although not directly related to the distribution of light and shade, color-changes do relate to the three-dimensional character of folds, and can thus serve as a form of modelling. The question is, whether they were adopted, as modern scholars have suggested, because of their formal capacity to model without value-contrast, and hence maintain a high key throughout a particular form, or whether they bore the connotations that derive from representing a particular sort of fabric. Shearman, for example, has argued that Andrea del Sarto used color-changes ‘of an entirely new order of subtlety’. Unlike those of the Quattrocento, which make a sharp contrast of chromatic and tonal value, from yellow to red or green to rose… they [del Sarto’s color-changes] move between values that are deliberately selected for their close association. Highlight and shadow are not, to a greater or lesser extent, made from separate pigments, but are carefully adjusted mixtures; cream-grey and lilac-grey may be coupled together, or shell-pink and lavender, turquoise and grey-green. When the color-change is a matter of nuance, like these, it can appear to be the fall of light on a lively and uniformly colored material.

Very little is known about the early history of shot fabrics; none has apparently survived from the Middle Ages or the Renaissance, perhaps because they were not figured and were thus less valued, and perhaps for the same reason very few were mentioned in the early sources. The earliest literary references seem to be in early fourteenth-century commentaries on the Sentences of Peter Lombard and the earliest inventory reference I have discovered is at Assisi in 1338.  The related discussion of lightening or darkening the colors of drapery in manuscript painting occurs in treatises from the mid-thirteenth century onwards;  but here it is often difficult to separate the idea of value from the idea of green, for example, so often encountered together the drapery of Trecento panting, had been since Antiquity regarded as the light or dark species of the same genus of hue. The pairing of black and blue, mentioned by Cennini in his chapter block-printing, and recognizable in many paintings, is subject to the modern eye to the similar confusions. Cennini (ch. LXXVII) assumed that cangianti draperies were suitable for angels, and this is often, though not exclusively, where they appear in Italian Trecento and Quattrocento painting. But what was perhaps most important that they clearly connoted silk, probably exotic silk, and hence great expense at the end of the sixteenth century Lomazzo, provides the most extensive treatment of cangianti combinations (Tratto della pittura Ill, X) and regards them as appropriate to nymphs and angels, insists that they are silka, and seeks to restrict the vast range of color-possibilities to those giving a convincing rendering of actual stuffs.