Were the primary colors we learned in school wrong?

Short answer

Not exactly wrong, but an incomplete and outdated simplification. Red, yellow, and blue are a useful starting palette that has served painters for centuries; but screens that work with light use red, green, and blue, and presses that work with ink use cyan, magenta, and yellow. There's no single "real" set of primary colors: primary colors are starting points chosen according to which material you're mixing and for what purpose.

Primary, secondary, and tertiary colors

The arrangement taught in school is simple. There are three primary colors, and every other color is derived from them. Mixing two primary colors gives a secondary color; mixing a primary color with the secondary color next to it gives a tertiary color. In the red, yellow, blue system, orange, green, and purple are the secondary colors; shades like red-orange or blue-green are tertiary.

This arrangement is useful as a classification: it's an easy way to build a color wheel, name colors, and learn to mix. The trouble starts when the word "primary" gets understood as a law of nature. The sentence "red, yellow, and blue make every color" doesn't turn out to be fully true even in the paint box a child is holding.

Where did RYB come from?

The idea of treating red, yellow, and blue as primary colors grew out of the long experience of painters and printers. In the early 18th century, the Frankfurt-born painter and engraver Jacob Christoph Le Blon tried producing color pictures by printing three separate plates, in blue, yellow, and red ink, one on top of another. In 1719 he patented this method in London, and in a small pamphlet published in 1725, he argued that painting could represent every visible object using yellow, red, and blue.

A few decades later, the English engraver and entomologist Moses Harris published color wheels that put red, yellow, and blue at the center as "fundamental" colors. The book's date varies in sources between about 1766 and 1770. Harris showed that mixing, lightening, and darkening these three colors could produce hundreds of colors. This wheel and its successors became embedded in the foundation of art education; the detailed history of the color wheel is the subject of DT-04.

What's interesting is that Le Blon's own practice was more flexible than his theory: in some prints he used a fourth plate in addition to the three. The idea that three colors alone don't do every job survives today in the black ink of printing; why that ink gets added is the question EK-08 takes up.

This idea had solid logic for its time. The pigments available to painters were limited, and there was no pure magenta or bright cyan pigment. Red, yellow, and blue were the trio that gave the widest range with the materials at hand.

The limits of RYB

To see the limits of RYB, you only need to go back to the paint box. When you mix red and blue, most of the time you don't get a vivid purple but a dark, muddy one. If you try to make a vivid pink or magenta with red, or a bright cyan with blue, you'll fail there too. The reason is the rule of subtractive mixing explained in DT-02: every paint subtracts a share from the light. Red paint already absorbs blue and green, and blue paint absorbs red and some green. When you mix the two, very little light is left, and the purple turns dark.

The true primary colors of subtractive mixing are the colors that each subtract only one-third of the spectrum: cyan absorbs only red, magenta absorbs only green, yellow absorbs only blue. When you mix any two of this trio, one primary color of light is left over: magenta with yellow gives red, yellow with cyan gives green, cyan with magenta gives blue. Seen this way, the "red" in the RYB system is actually magenta with a little yellow added to it; it isn't a primary color, it's a mixture.

This is the answer to the question on the lesson card: you can't get a vivid magenta from red, yellow, and blue, because magenta is a "purer" subtractor than all three. Magenta absorbs only green and leaves red and blue. When you try to make it from red and blue, you can't get back the blue that red absorbs along with green, or the red that blue absorbs; the mixture only gives a dark purple-pink. You can't make a color purer through mixing; you can only make it less pure.

There is no "real" primary color

So which trio is the right one? The question is being asked the wrong way. On screens that work with light, red, green, and blue give the widest range; in presses that work with ink, cyan, magenta, and yellow do. But no trio can produce the full range of colors the eye can see. The colors you can get by mixing three primary colors stay inside a triangle on a color map, and some portion of the visible colors always falls outside that triangle. How to read this map is the subject of BI-18, and the colors devices can't display are the subject of EK-04.

That's why experienced painters use a warm and a cool version of each color instead of a single red, a single yellow, and a single blue: a red leaning toward blue and a red leaning toward yellow, for example. When they want a clean purple, they pick the red that leans toward blue; when they want a clean orange, they pick the red that leans toward yellow. Primary colors aren't a rule; they're a toolbox.

Example — the purple in the classroom and skin tone in a portrait

In an elementary school art class, children are told to mix red and blue to make purple, and many of them end up with a muddy, brownish purple. The child hasn't done anything wrong; the absorption bands of the red and blue in the paint box overlap and leave very little light behind. If the same child were given a magenta and a cyan instead, they would get a far more vivid purple.

Skin tones also show the consequences of which primary colors you choose. A portrait painter can mix different skin tones from a limited palette of red, yellow, white, and black; but which red and which yellow are chosen for each skin changes the result directly. A cool red gives a pinkish undertone, a warm red an orangish one. No skin tone is produced by a single "correct" mixture; the method for producing skin tones with a limited palette is covered in SA-09.

Common misconception

Summary

1. The primary, secondary, and tertiary color arrangement is a useful classification, but "primary color" isn't a law of nature — it's a choice made according to the mixing system.
2. The red, yellow, and blue system became established in the 18th century through practitioners like Le Blon and Harris; RGB gives a wider range for light, and CMY for ink.
3. You can't make a color purer through mixing; that's why you can't get a vivid magenta or cyan from red, yellow, and blue, and no trio can produce every color.

Test yourself

1. (Easy) Which colors are the secondary colors in the RYB system?
2. (Easy) Which primary colors are used in printing?
3. (Medium) Why can't you get a vivid magenta from red, yellow, and blue?
4. (Medium) Who are the two 18th-century practitioners who put forward red, yellow, and blue as primary colors?
5. (Hard) Why does mixing magenta and yellow give red in subtractive mixing?
6. (Hard) What does a painter keeping two different reds on the same palette tell you about the idea of a primary color?

Answers

1. Orange, green, and purple; they're obtained by mixing two primary colors ("Primary, secondary, and tertiary colors").
2. Cyan, magenta, and yellow; adding black turns it into CMYK ("The limits of RYB").
3. Magenta is a pure subtractor that absorbs only green and leaves red and blue. Because red paint also absorbs blue and green, and blue paint also absorbs red, the light lost in the mixture can't be recovered; you can't make a color purer through mixing, so the result is only a dark purple-pink ("The limits of RYB").
4. Jacob Christoph Le Blon, with the Coloritto he published in 1725; Moses Harris, with color wheels that put red, yellow, and blue at the center ("Where did RYB come from?").
5. Magenta absorbs green, and yellow absorbs blue; when the two together subtract green and blue, only red light is left ("The limits of RYB").
6. It shows there's no single "primary red," that primary colors are tools chosen for a purpose: a red leaning toward blue is chosen for clean purples, and a red leaning toward yellow for clean oranges ("There is no 'real' primary color").

Sources

• Thinking 3D (Oxford), "Ink and light — Le Blon's Coloritto and Newton's Opticks."
• Lowengard, S., The Creation of Color in Eighteenth-Century Europe, Columbia University Press, Gutenberg-e, "Jacob Christoph Le Blon's Systems of Three-Color Printing and Weaving."
• Royal Academy of Arts collection, Moses Harris, "The Natural System of Colours."
• Colour Literacy Project, account of subtractive mixing ("Mixing processes").

Next step

DT-04 takes this lesson's wheel into its history: it covers how color wheels changed from Newton to Goethe, from Runge to Chevreul and Itten, and why colors get arranged on a circle instead of a straight line. BI-10, meanwhile, opens up the eye's side of the question: it covers why three stimuli can produce so many colors through the 19th century's color-matching experiments, what the theory fails to explain, and why screens use exactly three colors.