Thank you to Brilliant for sponsoring this whole week of SciShow! [ ♪ Intro ].

Humans have been talking to each other for hundreds of thousands of years. And for the most part, researchers have assumed we've all been making the same basic sounds, no matter what language we're speaking.

We're all working with the same equipment, after all. But a new study says that that is not the case. It claims that back before we developed agriculture, our teeth and jaws were different because we were chewing tough, fibrous foods, and that meant some sounds were harder to make.

Let's just say early hunter-gatherers weren't going around saying “fee fi fo fum” or “veni, vedi, vici.” According to this new research, humans have only recently made “f” and “v” sounds since they started eating softer foods, so in the last 10,000 years or so. This actually isn't a new idea. The hypothesis was first proposed by linguist Charles Hockett in 1985.

You see, scientists have generally assumed that the language-producing toolbox of humans hasn't really changed for about half a million years. As a result, you should expect to find roughly the same sounds in all languages. The easier a sound is to produce, the more often it should pop up.

According to linguists, the differences between the thousands of world languages mostly resulted from things like tiny pronunciation mistakes that spread culturally. Differences could even relate to geography, as certain sounds are easier to make at higher elevations. But Charles Hockett put one more explanation on the table: diet.

He noticed that the languages of modern hunter-gatherer societies lacked labiodentals. Those are sounds produced by touching or nearly touching the bottom lip to the upper teeth, like consonants like f and v in English. That led him to think differences between languages might actually come from how the food we eat affects the way we bite.

Both hominid fossils and modern hunter-gatherers have what he called an edge bite, where the top teeth lay directly on top of the bottom teeth, touching edge-to-edge. In childhood, humans start out with their top teeth overlapping their bottom teeth in a small overbite, or what Hockett called a scissors bite. But over time, tough foods wear teeth down, and to compensate for that loss, the teeth start to drift inward, leading to an edge bite.

And if your top and bottom teeth align perfectly, it's harder to get your bottom lip into the position where “f” or a “v” sound is possible. But according to Hockett, agriculture changed all that. We started eating less meat and more grains, and cooking came into the picture.

Softer food meant less wear and therefore helped humans keep their childlike overbite into adulthood. Scientists at the time didn't buy Hockett's hypothesis. They didn't think that wear and tear could make that big of a difference to the human bite.

Plus, archaeological evidence back then said that the timeline didn't add up: overbites appeared much later than agriculture did. In the end, Hockett backtracked on the idea, agreeing that the timing was an issue. But research since has shown that wear can change a person's bite dramatically, and it's now well established that overbites became more common after the rise of agriculture in the Neolithic era.

But a change in bite doesn't necessarily mean a change in language. You would need more research to make that connection. So that's exactly what an international team did for the new study published last week in the journal Science.

First, they created a 3D simulation of the human mouth and jaw to see if an overbite really did make labiodentals easier. It did: it takes about 29% less energy to make an “f” sound with an overbite than it does with an edge bite. Next, they created a statistical model that showed that modern hunter-gatherer societies only had about 27% as many labiodentals in their languages as agricultural societies.

And historically, they found that the statistical likelihood of a language having labiodentals increased over time, starting several thousand years ago, around the same time that things like dairy and cultivated grains became popular. And they became really common about 2500 years ago, when industrial milling became a thing in places like Europe. They also looked at how these sounds appeared in those languages.

Because a lot of language changes happen by accident, they figured that labiodentals would end up replacing sounds that require a similar mouth position. They put their money on bilabials, sounds like “p” and “m” that require you to bring your lips together. If a culture began speaking their language with an edge bite and then developed an overbite over time, it's understandable that a “p” sound might start to become an “f” sound here and there.

Sure enough, they found a lot of labiodentals where bilabials had once been. Like, the Italian “p” is related to the English “f”, which in part explains how “padre” became “father.” Not everyone embraces these findings, though. Some of the objections echo those from Hockett's time, saying that the researchers might be overstating how much diet can really affect language.

Others say that it relies on a lot of untested assumptions, including the idea that the use of agriculture is a good shorthand for a society's diet. But the biggest elephant in this room is the fact that tracing language differences to physical differences often leads to ethnocentrism or flat-out racism, something linguistics has had trouble with in the past. That's not a reason to abandon the results entirely, just a reason to treat them with extra care.

The researchers stand by their data, though, and say their study shows that everything humans do happens in the context of how they live. They've reiterated that they're not claiming that physiology is the only predictor of language, they just think it should get more credit, like it does for other human traits. Basically, there's no reason language and speech should be an exception.

In this study, the researchers used statistical models to show that diet probably influences the sounds we make. And if you want to really understand the math that led them to that conclusion, you might like this course on Probability from Brilliant.org. The course takes you from rolling dice to modeling the weather, all the while explaining the math that lets us determine whether an outcome is ‘likely'.

If you want to learn something quick every day, you might try your hand at their Daily Challenges. Every day, Brilliant puts out a set of problems to challenge your brain. And to solve them, you really have to apply your knowledge of things like statistics, logic, and science.

And right now, the first 200 people to go sign up at Brilliant.org/SciShow will get 20% off the annual Premium subscription, so you can view all of the Daily Challenges in the archives and unlock their dozens of problem-solving courses. You can learn, test yourself, and support SciShow all at the same time. That sounds pretty great! [ ♪ Outro ].