fb-pixel Skip to main content

Roses are red; violets are — red? How color terms arise

Zffoto - Fotolia

Since the turn of the 20th century, scientists have examined how humans around the world name colors in an attempt to answer one question: Does our language shape our worldview, or does our worldview shape our language?

Hannah Haynie, a postdoctoral associate at Colorado State University, teamed up with Yale University linguist Claire Bowern to find out. Their study analyzed a sort of evolutionary tree built from massive data found in field notes, dictionaries, and 20th-century records. The tree visualizes how color names potentially changed over time in the Pama-Nyungan language family, a group of indigenous Australian languages dating as far back as 6,000 years.


“It’s just like how, if you look at genes in people, you can look back at how they were transmitted along a tree,” Haynie said. “This brings a bunch of different sciences together to look at how language, our minds, and our world interacts together.”

The study tests a longstanding system of “basic color terms” called the Berlin-Kay theory, which came to light when scientists Brent Berlin and Paul Kay published a book in 1969 claiming that the way we divide up the visible color spectrum is hard-wired. The theory has two main parts: Only a few color name combinations show up in the world’s languages, and any new color names will always arrive in the same hierarchical order.

A stage-one language, for example, has two colors — black and white (dark and light). Languages that develop more color names will add red next, then green or yellow, and so on — all in a predetermined, universal order.

Normally, you’d think that colors are just something that are out there in the world, right? No matter who you are, that thing over there is red,” said Simon Greenhill, a senior scientist at the Max Planck Institute for the Science of Human History. “Berlin and Kay’s stuff says, ‘That’s not the case — not all languages have categorized these things, and not all languages care about these distinctions.’ ”

But this study reveals a few exceptions to that theory. For one, a language in the Pama-Nyungan family called Kukatj named four colors, but none of them are red. The phylogenetic tree also shows that another language, Wayilwan, may have lost blue as a color term over time, which goes against the idea that there’s a universal way that these words develop.


“It suggests that at least some of the steps of Berlin and Kay’s hierarchy might need to be looked at more closely,” Greenhill said. “Cultures might also be determining how people divide up or categorize the world into different colors.”

But even though Berlin and Kay’s theory isn’t perfect, most of the Pama-Nyungan languages still follow its hierarchy.

“Perhaps it’s not our language shaping our world,” Haynie said. “Perhaps it’s our cognition and some of its restraints that influence our language — world facts, like the finite spectrum of color or the finite terms we use to describe that.”

In other words, our prose is still likely at the mercy of how our brains visually process the world.

Kelly Kasulis is a journalist living in Boston and the deputy digital editor of The GroundTruth Project. Follow her on Twitter @KasulisK.