Scientists explain why humans universally perceive four primary colors

For more than a century, color scientists have been stuck on a puzzle that seems almost too simple to be hard: Why do humans everywhere see red, yellow, green, and blue as the fundamental colors of the world, while treating orange as a blend of red and yellow, or purple as a mix of red and blue? The question has haunted researchers because the answer doesn't match what we know about how our eyes actually work. We have three types of color-detecting cells in our retinas, not four. Our brains process color through neural networks that don't obviously map onto these four hues. Yet linguists documenting languages across the globe—written and unwritten alike—have found that cultures consistently organize their color vocabularies around these same four primaries, suggesting something deeper than culture or language is at work.

Alexander Belsten and Bruno Olshausen at UC Berkeley believe they've finally cracked it. Their theory, published in July in the Journal of the Optical Society of America A, proposes that the human brain has evolved to represent the natural world's colors in the simplest possible way: as combinations of just four opposing pairs. Red opposes green. Blue opposes yellow. This opposition is key. You can describe a color as reddish-yellow or greenish-blue, but the idea of a reddish-green or bluish-yellow feels nonsensical to us—because our brains don't encode the world that way.

The breakthrough came from examining what colors actually exist in nature. Olshausen, who studies how the visual environment shapes perception, asked Belsten to analyze massive datasets of natural scenes—millions upon millions of pixels from satellite imagery and calibrated photographs of landscapes, forests, and skies. What they found was striking: the natural world's color palette is far more limited and lopsided than the full spectrum suggests. Most pixels are colorless gray. The eye's sensitivity to natural colors doesn't spread evenly across the rainbow but clusters in three regions: red most prominently, then yellow-green, then blue-green. Crucially, except for red, nearly all the colors in nature are pale and unsaturated—think of a pale blue sky or the light yellow-green of vegetation, contrasted with the deeper reds of soil and tree bark.

This asymmetry is the key to understanding why four colors matter more than three. The researchers applied a principle called sparse coding, a concept developed decades ago by Berkeley neuroscientist Horace Barlow. The idea is that the brain encodes sensory information as efficiently as possible, using the fewest active neurons needed to represent the world. When Belsten modeled how neurons might represent the natural color distribution most efficiently, three types of neurons could theoretically do the job—one for red, one for yellow-green, one for blue-green. But four neurons are better. Four allow the brain to represent any natural color as a combination of just two of them, while also creating those mutually exclusive pairs: red versus green, blue versus yellow. This opposition gives the brain a parsimonious way to think about color that mirrors how we actually experience it.

Olshausen describes it like cardinal directions. If you say you're heading north, you're definitively not heading south. The same logic applies to color: red is not green, blue is not yellow. This opponent-process theory actually dates back to 1878, when German scientist Ewald Hering proposed that color vision worked through opposing pairs and called these four hues "urfarben"—primitive colors. But Hering's theory seemed to conflict with later discoveries about the three types of cone cells in our eyes, a model championed by Hermann von Helmholtz in the 19th century. For generations, researchers couldn't reconcile the two. Belsten and Olshausen's work suggests both were right: Helmholtz correctly described the physiology of the retina, while Hering correctly described the psychology of how we experience color—and that psychology is grounded in the actual structure of the natural visual world.

The implications extend beyond color theory. If the brain uses sparse coding to efficiently represent nature's colors, it likely uses the same principle for other sensory information—sound, touch, smell, even grayscale images. And if our primate relatives, which also have three types of color-detecting cones, face the same natural visual environment, they probably perceive colors the same way we do: as four fundamental, opponent hues. What Olshausen calls "almost too good to be true" is that all the pieces finally fit together—the physics of light, the biology of the eye, the structure of nature, and the universal human experience of color all point to the same elegant solution.