Science

What the colour wheel leaves out

Moses Harris's prismatic colour wheel with red, yellow and blue set evenly around the circle

The CIE 1931 chromaticity diagram, a horseshoe of colours with spectral wavelengths around its edge

Isaac Newton's colour circle from Opticks, divided into seven unequal sectors

Narcissus, a colour mezzotint by Jacob Christoph Le Blon printed from separate blue, yellow and red plates

A physical Munsell colour tree with painted chips arranged by hue, value and chroma

The colour wheel taught in schools grew out of Isaac Newton's 1704 circle and an 18th century painters' tradition of red, yellow and blue. This is where it came from, and what later colour scientists and artists found it could not show.

Newton's circle

The colour wheel most people meet at school is a blend of two separate histories. The first starts with Isaac Newton. In Opticks, published in London in 1704, Newton took the band of colours that a prism splits out of sunlight and bent it round into a circle, so that the red end met the violet end.¹²

Newton's circle was not divided evenly. He gave it seven sectors, red, orange, yellow, green, blue, indigo and violet, and sized them unequally to match the intervals of a musical scale, because he suspected that colour and sound shared the same harmonic proportions.¹³ Musicians and scientists in the 18th century took the idea up with some enthusiasm, but it was later discredited.¹ It is also worth knowing what Newton's circle did not contain. He described seven spectral colours of light and never proposed a theory of three primaries.⁴

The painters' primaries

The red, yellow and blue of the classroom wheel come from paint and print. In 1725 the German painter Jacob Christoph Le Blon, then working in London, published Coloritto, which named yellow, red and blue as the "Primitive" colours from which all others could be mixed.⁴ Le Blon also used those three colours to print pictures, laying transparent blue, yellow and red inks from three separate mezzotint plates on top of one another. He mentioned Newton's Opticks, but kept the two subjects apart, saying he was "only speaking of Material Colours, or those used by Painters".⁴ The later story that he built his printing method on Newton comes largely from a rival printmaker, Jacques-Fabien Gautier-Dagoty, who claimed Le Blon had followed Newton blindly.⁴

The wheel itself took its familiar form a few decades later. Moses Harris, who also published a book on English insects, set out red, yellow and blue at equal distances around a circle in The Natural System of Colours, a short guide for artists dedicated to Sir Joshua Reynolds.⁵⁶ His "prismatic" wheel has 18 numbered sectors built from those three, and a second "compound" wheel is built on orange, green and purple.⁶ The book carries no date. It is usually given as 1766, but the Royal Academy points out that Reynolds was only knighted in 1769, so it probably appeared between 1769 and 1776.⁵

Three sets of primaries, three sets of opposites

Red, yellow and blue work reasonably well for mixing paint, but they belong to one system among several. Paint and ink mix subtractively. Each pigment absorbs part of the light that falls on it, so every pigment added takes more of the light away.⁷ Coloured light mixes additively, and its primaries are red, green and blue.⁷ Shine red and green light onto the same white patch and the result is yellow, which surprises almost everyone who learned colour from paint first.⁸ The eye has three kinds of cone and behaves as though their signals are added together, which is why a screen can show its whole range of colours with red, green and blue dots.⁷ Colour printing uses a third set, cyan, magenta and yellow, usually with black added.³

The pairs of opposites, the complementary colours, move around depending on which system is in use. On the painter's wheel red sits opposite green and blue sits opposite orange. With light, the complement of red is cyan and the complement of blue is yellow.⁹ Newton's own circle paired red with a blue close to what would now be called cyan.⁹ The visual system adds another layer. The physiologist Ewald Hering proposed that colour vision is organised around opposed pairs, red against green and blue against yellow, and cells that respond in exactly that push and pull way have since been found in the retina and in the pathway from the eye to the brain.⁸ The same wiring explains afterimages. Stare at a green patch for a minute, look at a white wall, and a reddish ghost appears.⁸

Colour in three dimensions

A flat circle records hue, the quality that separates red from yellow from blue. Colours also differ in how light or dark they are and in how intense they are. The American painter Albert Munsell, who taught at the Massachusetts Normal Art School for 37 years, named these value and chroma, and made them, with hue, the basis of a measured colour system.¹⁰ His first book, A Color Notation (1905), proposed fitting every colour into a sphere. He later abandoned the sphere because colours varied too much in brightness to fit inside it, and arranged them instead in what he called a colour tree. His colour atlas followed in 1915, and a posthumous version, the Munsell Book of Color of 1929, is still in use.¹⁰

The 20th century brought more systems. In 1931 the International Commission on Illumination (CIE) adopted a mathematical "standard observer" built from experiments run in England in the late 1920s by William David Wright, who tested about ten people, and John Guild, who tested about seven. Each observer adjusted a mix of red, green and blue light until it matched a test colour.¹¹¹² The CIE XYZ space derived from that work is still the reference against which many other colour spaces are defined, so a good deal of modern colour management rests on the eyes of those 17 people.¹¹¹²

In Sweden, Anders Hård, Lars Sivik and Gunnar Tonnquist developed the Natural Colour System from 1964. Following Hering, it describes every colour by how closely it resembles six elementary colours, white, black, red, yellow, green and blue, and it is concerned with how colours look rather than how they are mixed.¹³¹⁴ It became the Swedish reference standard for colour notation in 1979.¹³

Albers and context

Josef Albers came at the problem from the studio. He taught at the Bauhaus as a master from 1925 until the Nazis closed the school in 1933, then at Black Mountain College in North Carolina, and in 1950 became chairman of the design department at Yale.¹⁵ His Interaction of Color, published by Yale in 1963 as a portfolio of 81 screenprints with a volume of text, was meant as a teaching aid.¹⁶¹⁵ Its plates use colour illusions to show how much a colour changes with whatever surrounds it. Albers called colour "the most relative medium in art".¹⁵

A fixed wheel suggests that colours sit in fixed relationships to one another. Albers spent his teaching life showing students that the same patch of paint can be pushed warmer, cooler, lighter or darker by its neighbours, and Munsell, the CIE and the Natural Colour System each map parts of colour that a circle of hues has no room for. The wheel still works as a quick map of hue. Most of what happens on a canvas sits outside it.