Aref Karimi / Stringer / Getty

Thousands of years ago, small groups of humans across the globe began to transition from hunting and gathering their food to raising and planting it instead. They milked cattle, milled grains to make soft bread, and used new inventions like pottery to preserve meat and vegetables. And once they did that, they could start spicing up their speech by throwing some f and v sounds into the mix.

At least, that’s according to a new study published Thursday in the journal Science. The authors argue that sounds like f and v weren’t part of human language until farming appeared during the Neolithic age. Agriculture, they say, allowed humans to eat soft foods, which changed the way their jaws developed throughout life, which shaped the kinds of sounds their mouths were capable of making.

This shift would be an exception to a core rule of linguistics, called the uniformitarian principle, which posits that humans’ ability to use language has not significantly changed since language itself first appeared. “Basically, the uniformitarian principle is necessary to do historical linguistics,” Anthony Yates, a linguist at UCLA, told me. It’s hard to say when exactly humans started speaking, but most estimates place the date at least 50,000 years ago. Agriculture, meanwhile, sprung up during the Neolithic, some 12,000 years ago. The idea that humans weren’t using f’s and v’s for the first 38,000 years of our linguistic history is a striking rebuke to uniformitarianism.

Linguists spend a lot of time thinking about all the strange ways humans can shape their mouths, and how that affects the sounds they can make. Anthropologists believe that since at least the Neolithic, humans have been born with overbites (in which the top teeth come down over the bottom teeth) and overjets (in which the top teeth are positioned more forward than the bottom teeth). If a person’s teeth are exposed to a lot of wear and tear over her lifetime, her jaws will shift to put her teeth in an edge-to-edge alignment instead. Extreme wear was common early in human history, when most of our ancestors were hunters and gatherers. In farming societies, on the other hand, people tend to put less strain on their teeth, and retain overbite and overjet for their whole life.

John Lukacs, an anthropologist at the University of Oregon who was not involved in the new study, told me that agricultural diets are easier on the teeth in two main ways. First, they’re less diverse, made up mostly of staple grains such as corn, wheat, and rice. They’re also moderated by food-processing technologies such as grain mills and pickling jars. “It’s not just what you eat,” Lukacs said. “It’s how you prepare what you eat.”

Mouths with overbites and mouths with teeth lined up edge-to-edge are good at different things. F’s and v’s belong to a group of sounds called labiodentals because they’re made by touching the upper teeth to the lower lip, and the shape of a person’s mouth determines how easy they are to produce. If you’re reading this article, odds are you live in an agricultural, overbite-bearing society. When you say the word foot, you don’t have to move your mouth too much from its resting position, because your top teeth are already lined up with your bottom lip to make the labiodental f sound. If you move your teeth into an edge-to-edge position, roughly the way an adult hunter-gatherer’s jaws would be arranged, making an f sound involves a lot more movement. (Try it.) In fact, the authors of the Science study estimated that it’s nearly 30 percent easier for people with an overbite to make labiodental sounds than it is for people with edge-to-edge bites.

Because it’s so difficult for people with edge-to-edge bites to produce sounds like f and v, the study’s authors figured they would be unlikely to say them by accident, or to incorporate them into their languages. They checked to see whether they could find this pattern playing out in the real world by comparing the sound systems of languages across the world with the subsistence style of the people who speak those languages. About half of the world’s languages use labiodental sounds, but on average, languages spoken by hunter-gatherer societies turned out to use fewer than one-third the number of labiodental sounds as their agricultural counterparts.

Noreen von Cramon-Taubadel, an anthropologist at the University at Buffalo, called the study “really exciting” because it crossed disciplinary boundaries. “We often tend to separate out biological factors from cultural factors,” she told me, but “in humans, it’s really hard to do that.”

The authors of the study are well aware of that tension. In a call with reporters, one of the authors, Balthasar Bickel, lamented that linguistics is often grouped with subjects such as literature and art, but scarcely (if ever) mentioned in biology classrooms. This has enabled what Bickel, a linguist at the University of Zurich, called an “introvert perspective” in the field. “Traditionally,” he said, “change and diversification in language has been chiefly seen as driven by internal processes”—that is, psychology, not biology. Bickel thinks that focus is holding linguists back. “I believe there is a huge potential out there for studying language as part of the biological system it really is embedded in,” he said.

Mixing biology and psychology doesn’t always lead to scientific enlightenment. In the late 19th century, eugenicists were eager to insist that biological factors are the primary driver of humans’ mental abilities, which, they thought, justified keeping those deemed undesirable from reproducing. That concept is known as biological determinism, and on first glance, it’s possible to catch a whiff of it in this study. After all, the authors are arguing that a biological fact (jaw shape) affects the kind of language a person can use (a psychological process). But none of the researchers I spoke with seemed particularly concerned that biological determinism is at play in this study.

Lukacs, the anthropologist, said that the paper provides evidence of a “biological linkage” or “covarying function” between diet and speech, but it does not argue that diet directly causes or determines how a person speaks. Yates, the linguist, said he isn’t worried about the study sparking a deterministic renaissance in linguistics, because it showcases a “very minor, principled exception” to the uniformitarian principle. “I don’t think they’re going to find more,” he told me. “This is not something that we should have to spend a huge amount of time worrying about.”

That’s a sharp contrast to the authors of the study, who said on the call with reporters that they hope more researchers will follow their lead and investigate whether other classes of sounds, besides labiodentals, might have appeared relatively late in humans’ linguistic history. On a later call, Bickel told me he’s especially interested in the farming cultures of Mesoamerica, a region that received somewhat less attention in the new study. But that doesn’t mean he’s plotting some kind of biodeterministic takeover.

“It’s very clear and also very important for us to note that the mechanisms that we found are not deterministic; they are probabilistic,” Bickel said. In other words, just because agricultural societies are more likely to use f’s and v’s, they aren’t necessarily resigned to that fate. Japan, for example, has a strong and long-standing agricultural tradition, but the Japanese language doesn’t use any labiodentals.

The study pokes at some of the most fundamental seams of linguistics: its somewhat ill-fitting niche between literature and cognitive science, its attempts to reconcile the universality of language abilities between humans with the enormous diversity of individual sound and syntax systems worldwide.

Bickel told me he thinks the new findings pull at the very fabric of Western philosophy, too. His colleagues in the humanities, he said, often talk about human behavior as if it’s completely “isolated from nature” because they’re “uncomfortable even comparing” it with the actions of animals. If we want to make “any progress” toward understanding the very thing that makes us human, he said, we need “to understand language as part of the evolutionary history that our species went through.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.