Does color really exist?

Short answer

Color does not live only in the light, only in the object, or only in the brain; it appears only when the three meet. Light is a colorless stream of energy made of different wavelengths, the object holds onto part of that stream and sends the rest back, and the brain turns the returning mixture into an experience of color. When one of these three links breaks, the color we see either changes or disappears.

Light itself is a colorless wave

Sunlight entering a room is a mixture of electromagnetic waves that differ from one another only in wavelength. Our eyes are affected by only a very narrow slice of these waves: roughly 380 to 780 nanometers. Waves at the short end of this slice appear to us as blue and violet, and those at the long end appear as red. But the phrase "appear as" matters, because the wave itself carries no dye, no pigment belonging to blue or red. Wavelength is a length; just as the number of vibrations of a guitar string is not itself a sound, a length on its own is not a color.

Isaac Newton was one of the first to state this idea clearly. In Opticks, published in 1704, he writes that calling rays "red" or "blue" is a crude way of speaking, and that the rays themselves aren't actually colored. In his view, light only has the power to produce a particular color sensation in the eye. Which experiments led Newton to this conclusion is a separate story (BI-03); what wavelength is and how far the spectrum stretches is the subject of BI-01.

The object selects the light: what the apple does

The skin of a red apple contains pigments called anthocyanins; the most dominant of these is cyanidin-3-galactoside. These molecules absorb most of the short- and medium-wavelength portion of the light that hits the apple, meaning they turn it into heat and hold onto it. What's left, reflected back, is mostly long-wavelength light. What we call the apple's "redness" is this very act of selection: which wavelengths it holds onto and which it lets go.

This leads to a very important conclusion: the object doesn't produce a new color, it only selects from the light that reaches it. If there's nothing to select from, no red appears either. The detail of how a surface absorbs and reflects light — that is, selective reflection — is explained in BI-04 through a tomato; why molecules hold onto certain wavelengths is the subject of MA-01.

Color gets its name in the brain

When the light reflected from the apple reaches your eye, the three types of cone cells in your retina respond to it in different degrees. The brain compares the ratios of these three responses and gives the result a name: red. In other words, the experience you live as "red" isn't born in the light itself, but in the outcome of this comparison. How cone cells work is explained in BI-09, and the parts of the eye in BI-08; the one thing you need to know here is this: color perception isn't a measurement, it's an interpretation.

The strongest evidence that it's an interpretation is that the brain takes context into account. The same apple sends your eye different mixtures of wavelengths in the morning sun and under an evening lamp, yet you see it as red in both cases. The brain predicts that the light has changed and corrects its calculation accordingly. How this correction works and where it goes wrong is covered in KE-03 and BI-13.

The same chain holds for human skin too. Skin color depends on how melanin and the hemoglobin in blood absorb light, and this absorption follows the same physical rules in light skin and in dark skin alike. What changes is which wavelengths come back, and in what proportion. The biological story of skin color is told separately in DO-06.

Example — a red car under sodium light

Low-pressure sodium lamps, once used in older highway tunnels and still found on some city streets, give off light at almost a single wavelength: an orange-yellow light around 589 nanometers. Under such a lamp, no color difference remains between a red car and a blue car; both look like nothing more than a lighter or darker gray-orange. The cars' paint hasn't changed, and neither has your brain. The only thing that's changed is that the light the paints can select from is now just a single wavelength.

The same thing happens with skin: under sodium light, the color difference between different skin tones is erased, leaving only a difference in lightness and darkness. This is why choosing street lighting isn't only a matter of energy; it also determines how people see one another.

Common misconception

Summary

1. Light is a colorless stream of energy made of different wavelengths; the visible range is roughly between 380 and 780 nm.
2. The object doesn't add color to light, it selects from the light that reaches it; its color is which wavelengths it absorbs and which it reflects.
3. Color perception is an interpretation that arises from comparing the eye's responses in the brain, and it's corrected according to context.

Test yourself

1. (Easy) Roughly what range is the wavelength of visible light?
2. (Easy) What is the dominant pigment group in a red apple's skin?
3. (Medium) Is an apple red in pitch darkness?
4. (Medium) Why do red and blue cars look the same color under a low-pressure sodium lamp?
5. (Hard) What did Newton mean when he wrote that the rays themselves aren't colored, and how does this idea connect to the brain's role in color perception?
6. (Hard) A student says, "The apple's color doesn't change, because its pigment stays the same." Separate what's true and what's false in this statement.

Answers

1. Roughly between 380 and 780 nanometers; the limits aren't fixed and shift depending on the observer ("Light itself is a colorless wave").
2. Anthocyanins; the most dominant is cyanidin-3-galactoside ("The object selects the light").
3. The apple's pigment is still there in the dark, meaning its tendency to reflect long wavelengths is preserved. But since there's no light to select from and no vision to interpret that light, the experience of red doesn't occur. If by "red" we mean the surface's property, yes; if we mean the color we see, no; the philosophical discussion of this distinction is in FE-01.
4. The lamp emits light at almost a single wavelength; neither paint can find another wavelength to select from, so only a difference in lightness remains ("Example — a red car under sodium light").
5. He meant that light itself has no color, only the power to produce a particular color sensation; that color itself arises in seeing. This connects to the idea that color perception is formed by the brain interpreting the eye's responses ("Color gets its name in the brain").
6. True part: the pigment, meaning the way it selects light, stays the same. False part: the color we see doesn't depend on pigment alone; when the light changes, the reflected mixture changes too, and color perception shifts accordingly, or the brain corrects for it ("Common misconception").

Sources

• Newton, I., Opticks, London, 1704; the definition following Book I, Part II, Proposition II. Online text: The Newton Project, University of Oxford.
• CIE S 017:2020, International Lighting Vocabulary (ILV), 2nd edition, the entry "visible radiation."
• Honda, C. and Moriya, S., "Anthocyanin Biosynthesis in Apple Fruit," The Horticulture Journal 87(3): 305–314, 2018.
• RP Photonics Encyclopedia, the entry "Sodium Vapor Lamps."

Next step

BI-01 opens up the first link of this lesson: it shows how narrow a slice of the electromagnetic spectrum visible light is, and why our eyes are tuned to exactly this range. BI-04 opens up the second link: it explains reflectance curves, transparent versus opaque material, and the difference between glossy and matte surfaces through a tomato; the question of a tomato under blue light is a continuation of the sodium lamp example.