The internet abounds with techniques for teaching elementary schoolers the difference between warm and cool colors—an often-invisible, somewhat flexible line down the middle of the color wheel to separate warm reds, oranges, yellows, and browns from cool blues, greens, purples, and grays. The balance between them is said to enhance the beauty of Baroque landscapes and the Mona Lisa. Interior designers claim that cool colors recede and make rooms expand, while warm colors make rooms cozier.

Still, the basis for the warm-cool divide has remained murky, largely resting on the sometimes ambiguous and overlapping feelings different colors give people, as opposed to any clear scientific distinction. But a new study might change that: Across languages, it suggests, warm and cool colors can be distinguished by how easy they are to communicate. When you’re trying to describe a color to someone else, that person will identify the correct one faster if it’s warm rather than cool. This result has implications for the evolution of color vision in humans and other primates, and even the reason language developed to begin with.

The study, which was published Monday in the Proceedings of the National Academy of Sciences, asked people who speak three different languages—American English, Bolivian Spanish, and an Amazonian language called Tsimane—to play a guessing game. The researchers gave one participant (the speaker) a randomly selected color chip, then had them describe it to another participant (the listener) using a single color term. The listener picked out all the chips that could fall into the category described by the term—anything and everything that looks like a yellow, or a pink, or a green, and so on.

The number of chips the listener picked out was the study’s proxy for “efficient” communication: The fewer chips that a word corresponds to, the fewer guesses it would take for the listener to ultimately find the right one with just the single color term as a clue.

In addition to the three language groups, the researchers used data from the World Color Survey, a publicly accessible collection of results from anthropologists performing similar guessing games among speakers of 110 languages around the world. When the researchers ranked the data from all 113 languages, a division between warm and cool colors started to take shape: Listeners selected fewer possible chips when a speaker described a warm-colored one than a cool-colored one.

The implication: If you were to take the spectrum of colors that are perceptibly different to humans and chop it in half, every language would have more words for describing the warm half than the cool half. And that would mean each word for a warm color refers to fewer colors than the cool words. The warm words are more specific—and more efficient at getting the point across.

“That made an incredible amount of sense,” says Bevil Conway, one of the paper’s coauthors and a researcher at the National Institutes of Health’s National Eye Institute. Conway’s previous research focused on primate vision, which had already clued him in to the fact that warm colors seem to be special: Fruit, for example, has an evolutionary pressure to attract primates so we can spread its seeds, and it is often red, pink, or orange-colored. The division in the new study has led the researchers to suspect that humans are fundamentally more interested in warm-colored things.

“We go to great trouble to make a name for something,” Conway says, so it’s reasonable to think that we’re more likely to name things we care more about. If we care more about warm-colored things, his thinking goes, we’ll probably name more warm colors, and leave cool colors with fewer specific words to describe them. You might have seen this play out in your daily life if you’re an American English speaker. One roommate will call the new rug blue, another will call it green; one middle-school dance attendee requests that their date match their green getup, and the date shows up in turquoise.

Still, says Paja Faudree, a linguistic anthropologist at Brown University, “it doesn’t seem necessarily obvious that that’s some kind of universal distinction” between warm and cool colors among cultures worldwide. Faudree, who wasn’t involved in the study, is skeptical about how reliable the data is, since “not all languages have color words that are abstract qualities,” separate from other aspects of an object: The grammar of a language, for instance, can be such that color terms are inextricable from words describing qualities like brittleness or inflammability. Conway acknowledges that most languages don’t have a single term for the concept of color, so it can be difficult to even explain the guessing game to many language groups.

Beyond words for color, linguists are fond of debating the reason that language exists at all. Noam Chomsky has famously argued that language didn’t originate for communicating ideas to other people, but so that early humans could develop advanced thought. Ted Gibson, a coauthor on the study from the Massachusetts Institute of Technology, says the results show that language must have evolved at least partially for the purpose of communication, since it’s optimized for effectively describing the things we care about. All the language people use is just a souped-up version of his experiment’s guessing game, he contends: “Communication consists of something I want to tell you and a code we both know. How many guesses is it going to take you to figure out what I meant given what I said?”

Conway, too, says the study addresses an existential question in vision science: the reason humans can see color in the first place. Plenty of vision scientists, Conway says, think that color vision is just “a little aftereffect.” But his results suggest that “there’s something really fundamental about the contribution of color to object vision”: our ability to look at the world around us and pick out the objects in it. That ability, Conway says, is too often taken for granted.