michel-eugène chevreul's "principles of color harmony and contrast"

The Principles of Harmony and Contrast of Colors by Michel-Eugène Chevreul – This classic "color theory" text, published in 1839 as The Law of Simultaneous Color Contrast (translated into English in 1854), is an artistic milestone, one of the first systematic studies of color perception and a compendium of color design principles that many 19th century French painters from Delacroix to Matisse attempted to apply in their art.

After an illustrious academic career studying fats and waxes, the chemist Michel-Eugène Chevreul (1786-1889) was appointed by royal decree to be director of dyes at the national Gobelins textile factory in Paris, where he worked for 28 years (1824-52) on chemical research and quality assurance in the dyes used for fine fabrics and textile designs. (He devoted much of his labor to developing more lightfast blue and violet dyes.) This middle position between organic chemistry, manufacturing technology and consumer response brought basic color problems to Chevreul's attention, in particular the apparent shift in the depth of black fabric depending on the colors surrounding it.

Through observation, experimental manipulation, and basic color demonstrations practiced on his coworkers and customers, Chevreul identified his fundamental "law" of the simultaneous contrast of colors:

"In the case where the eye sees at the same time two contiguous colors, they will appear as dissimilar as possible, both in their optical composition [hue] and in the height of their tone [mixture with white or black]."

He distinguished three situations in which this color contrast could be observed: simultaneous contrast, which appears in two colors viewed side by side (which was familiar to painters since the Renaissance), successive contrast, more commonly called negative afterimages (which had been studied since the mid 18th century), and mixed contrast, which appears in two colors viewed one after another (that is, the second color is mixed with the negative afterimage of the first color) — an effect Chreveul observed in the distorted color judgments of textile buyers who examined first many fabrics of one color, then of another.

Chevreul claimed to predict the visual effect of simultaneous contrast across all these situations with a single rule: if two color areas are seen close together in space or time, each will shift in hue and value as if the visual complementary color of the neighboring or preceding color were mixed with it. Thus, if a dark red and a light yellow are seen side by side, the red will shift as if mixed with the visual complement of light yellow (dark blue violet), while the yellow will shift as if mixed with the complement of the dark red (light blue green): the red will appear shifted toward violet, and the yellow toward green. (Similar shifts appear if either color is seen after the other.) At the same time, dull or near neutral colors will make saturated colors more intense, though Chevreul was not clear about this effect.

Chevreul noted that these apparent color shifts are strongest when the color areas are viewed side by side rather than far apart, are equal in size and not too large, and are viewed under subdued light. Most of his demonstrations use sheets of painted or dyed paper — the method adopted by Joseph Albers — as papers can be cut to any convenient size or shape and switched around to test all possible color combinations. (For more information on these color effects, see the sections on simultaneous color contrasts, full color harmonies and near neutrals and color design.)


comments : requests

Chevreul developed a hemispherical color model to explain and predict these various contrast effects. He adapted Newton's hue circle by locating the six subtractive "primary" (red, yellow and blue) and secondary (orange, green and violet) hues at equal sixths around the circumference, added the tertiaries (violet red, orange red, orange yellow, yellow green, green blue and blue violet) in between, just as Moses Harris had done. Finally Chevreul divided each twelfth into six intervals, for 72 hue increments in all. The mixing complement of each hue was placed directly opposite it on the hue circle.

Each hue increment was represented by a normal scale of tones (image, right) which presented the most saturated pigment (or pigment mixture) for that hue in a series of 22 value steps, from pure white at step 0 (the bottom of the scale, placed at the center of the hue circle) to pure black at step 21 step (at the circumference of the hue circle). The same numerical step of the normal scales was intended to have the same height (lightness) across all hues. This meant that the location of the "pure" hue within each scale depended on its lightness: light hues such as yellow would be closer to white, dark hues such as blue would be closer to black. All steps above the pure color were mixtures of the pigment with black only, all steps below it with white only.

To transform the hue circle of spokelike normal tone scales into a hemispherical color model, Chevreul specified that a vertical axis be constructed at the center of the hue circle. This axis would be the "normal scale of tones" for black — that is, a 22 step gray scale, with white at the bottom step 0 and black at the top step 21. Then each step of this normal black scale would be mixed with the corresponding step of the normal tone scale for each hue, in 9 equal visual steps according to the visual mixture ratios 9:1, 8:2, ... 2:8 and 1:9. These 9 new scales would be arrayed as a "fan" rising above the normal tone scale (image, below). When this operation was completed for all 72 normal tone hue scales, the result would be a color hemisphere, cloaked in black with white at the center.

chevreul's color model

the hue circle is exemplified by 72 normal scales of tones, arranged with white at the center and black at the circumference; as shown for yellow, each normal scale produced 9 broken scales of tones by means of increasing proportional mixtures with the achromatic gray scale located as the vertical axis of a hemisphere

Although (to my knowledge) no physical model of Chevreul's color hemisphere was built in his lifetime, editions of Chevreul's book included illustrations of the hue circle in large format, sumptuous lithograph color plates that were printing achievements for the time. This makes Chevreul's model one of the first mixture color models, because his color system had to be translated into explicit ink recipes, or printing procedures, in order to be reproduced.

There are several conceptual problems with Chevreul's model. In particular, his concept of "tone" confounds lightness and saturation. Mixture with white or black paint will always change a paint's lightness, though in opposite directions (lighter or darker). However, saturation remains relatively unchanged as an oil or acrylic paint is mixed with black, but saturation always decreases when it is mixed with white. As a result, Chevreul often does not distinguish between lightness and chroma in his analysis of color effects, and does not always describe simultaneous chroma shifts accurately or unambiguously. (The distinction between brightness or lightness and chroma was not clarified until the second half of the 19th century, in the writings of Hermann von Helmholtz and Ogden Rood.)

a "normal scale of tones"
for orange red and yellow

the pure pigment is marked by a white dot; the lightness of each step of the scale is the same across all 72 hues of the color hemisphere

Despite his reputation as a groundbreaker, Chevreul was actually a conservative thinker at the end of a long tradition. His analysis of chroma and lightness variations entirely in terms of mixtures of pure pigment with white or black goes back directly to the Della pittura of Leon Batista Alberti (1436) and was the method used by all painters to manipulate color mixtures. Complementary contrasts were already well known to artists of the time, and studious artists would have encountered Chevreul's key ideas as early as Alberti or the notebooks of Leonardo da Vinci (first published in French in 1651):

"Of colors of equal lightness, that will look brightest which is against the darkest background, and black will display itself at its darkest against a background of greatest whiteness. And red will look most fierce against the yellowest background, as do all colors surrounded by their directly contrary color."

And Chevreul certainly knew Aristotle's comments on the visual mixtures of different colored yarns. In any case, he asserted priority by exploring these historical topics with thorough experimentation.

Chevreul's book is a laborious survey of color effects using colored papers, yarns, glasses and afterimages; of colors placed against one another or in contrast with white, gray or black; of colors in printed or paper designs, in paintings, in clothing and textiles; and of the effects of different hues and intensities of illumination, including light through stained glass windows.

Chevreul summarized all this work in terms of six color harmonies, and again these are largely traditional:

• harmonies of scale, produced by the pure hue across a range of "tones" (mixtures with either white or black, such as a drawing in red, white and black conté crayon)

• harmonies of hue, produced by analogous hues, all within a narrow range of tone (that is, related hues of approximately the same lightness, or mixture with white or black)

• harmonies of a dominant colored light, produced by any selection of colors that mimicks the appearance of contrasting colors viewed under a colored light or through stained glass [a transmission filter] (i.e., colors are first chosen to produce desirable contrast, then all are mixed subtractively with an additional color)

• harmony of contrast of scale, produced by the contrast in a single hue at two widely contrasted "tones" (which may mean a contrast in lightness and/or in saturation)

• harmony of contrast of hues, produced by the contrast between analogous hues differing in "tone", and

• harmony of contrast of colors, produced by the contrast between complementary or near complementary hues, sometimes augmented by a contrast in tone — a contrast which Chevreul called "superior to every other" when the colors are of similar tones.

These harmonies or contrasts could be enhanced by combining with white or black backgrounds or borders, especially when the colors were "luminous" (saturated or light valued). Chevreul's prescriptions come at the end of the book, however, and few artists read that far because of the exhaustive and complex description of color research that preceeded it. His reputed wide influence on 19th century artists was actually channeled through the concise summary of Chevreul's work offered in the Grammar of the Graphic Arts (1867) by the French art critic Charles Blanc (1813-1882).

Inconsistent terminology and Chevreul's faulty understanding of color vision make many parts of the text either confusing or inaccurate. Chevreul claims, for example, that the "primary" colors are created by "pure" red, yellow and blue light, and that contiguous colors "take away colored rays" from each other. We now know that paint colors have little in common with light "colors", that the subtractive "primaries" must reflect a wide spectral range to be effective and are therefore not "pure" (yellow for example is a mixture of "red" and "green" light), and that the color shifts Chevreul described occur not in the light but entirely in the mind.

Artists can benefit from repeating Chevreul's color demonstrations in order to see the color effects for themselves — a study that is more compelling than just reading the book. Faber Birren's annotations at the back, read straight through, provide a reasonable outline of Chevreul's main points, although in the details Birren's comments are sometimes weirdly inaccurate. The plodding translation doesn't help, as it reproduces the 19th century French vocabulary and syntax almost word for word: the result isn't French but it's not really English either.