A ‘Mic Drop’ on a Theory of Language Evolution

An anatomical drawing of a human larynx
Getty / The Atlantic

Put your fingertips against your throat and say “abracadabra.” (Don’t whisper; it won’t work. Feign a phone call if you have to.) You should feel a buzzing—that’s your vocal folds vibrating inside your larynx.

The larynx, also called the voice box, is where the trouble begins: Its location is, or was, supposed to be the key to language. Scientists have agreed for a while that the organ is lower down the throat in humans than it is in any other primate, or was in our ancestors. And for decades, they thought that low-down larynx was a sort of secret ingredient to speech because it enabled its bearers to produce a variety of distinctive vowels, like the ones that make beet, bat, and boot sound like different words. That would mean that speech—and, therefore, language—couldn’t have evolved until the arrival of anatomically modern Homo sapiens about 200,000 years ago (or, per a fossil discovery from 2017, about 300,000 years ago). This line of thinking became known as laryngeal descent theory, or LDT.

A new review paper, published yesterday in Science Advances, aims to tear down the LDT completely. Its authors argue that the anatomical ingredients for speech were present in our ancestors much earlier than 200,000 years ago. In fact, they propose that the necessary equipment—specifically, the throat shape and motor control that produce distinguishable vowels—has been around as long as 27 million years, when humans and Old World monkeys (baboons, mandrills, and the like) last shared a common ancestor.

Recommended Reading

LDT “told people, basically, don’t bother to go look” for speech abilities in anything other than modern humans, says Thomas Sawallis, one of the authors of the new paper. Those speech abilities could include distinct vowels and consonants, syllables, or even syntax—all of which, according to LDT, should be impossible for any animal without a human vocal tract. “There was always this idea,” says Greg Hickok, a cognitive-science professor at the University of California at Irvine who was not involved in the study, “that there was one thing that had to happen and that released the linguistic abilities.” For Noam Chomsky and his followers, that thing was the invention of syntax. For proponents of LDT, it was the reshaping of the human throat.

Part of the reason LDT caught on to begin with is that language evolution, as a field, lacks concrete data. As John Locke, a linguistics professor at Lehman College, put it, “Motor control rots when you die.” Soft tissues like tongues and nerves and brains generally don’t fossilize; DNA sequencing is impossible past a few hundred thousand years; no one has yet found a diary or rap track recorded by a teenage Australopithecus. So the anatomical argument presented by LDT gave researchers something to latch on to. Until the ’60s, people who studied language evolution “were considered crackpots because they didn’t have any data,” Locke says. “When you’ve got nothing on the table, a little something goes a long ways.”

The researcher generally credited with developing laryngeal descent theory is Philip Lieberman, now a professor at Brown University. He called the new paper “just a complete misrepresentation of the entire field,” among other things. One of the quantitative models the new study relies on, he says, doesn’t properly represent the shape of the larynx, tongue, and other parts we use to talk: “It would convert a mailing tube into a human vocal tract.” And according to Lieberman, laryngeal descent theory “never claimed language was not possible” prior to the critical changes in our ancestors’ throat anatomy. “They’re trying to set up a straw man,” he said.

Yet other experts I spoke with told me that setting an upper bound on when speech, and therefore language, could have possibly evolved was exactly the effect that LDT had on anyone studying language evolution. Hickok said that when he was being trained in linguistics, “this was an established, almost dogmatic idea.” The new study is a dramatic reversal of the status quo, he said: “The phrase that came to mind when I finished it was mic drop.”

Still, he doesn’t agree entirely with Sawallis and his co-authors’ conclusions. Rather than 27 million years, Hickok proposes that the earliest bound on any sort of speech ability would be nearer to human ancestors’ split with the Pan genus, which includes chimpanzees and bonobos, our closest living relatives. That split happened about 5 million to 7 million years ago—certainly longer than 200,000 years, but a far cry from 27 million. Lieberman argues that the precursors of speech might have emerged about a little more than 3 million years ago, when artifacts like jewelry appear in the archaeological record. The idea is that both language and jewelry are intimately related to the evolution of symbolic thinking.

Locke, meanwhile, says that “not for a minute” does he believe that the dawn of speech could have happened 27 million years ago. That’s because he believes that speech and language couldn’t have evolved before humans began living in large social groups, which was likely a later development.

Given the dearth of hard data, “there really is almost no evidence for any of these positions,” Hickok said. But pushing back the 200,000-year limit “gives us a little bit more room to think about this stuff.” Too often, he said, scientists assume that the brain has always been a lean, mean learning machine, and “evolution kind of stops when you get to the head.” Focusing less on our throats could allow for more investigation into what about our brains made us human, and when. The downfall of the LDT also could lead to more research into signs of speech in nonhuman primates, Sawallis hopes, which would help pinpoint the beginnings of language a bit more precisely than sometime in the last 27 million years.

“There’s a bunch of stuff we haven’t found” in the study of language evolution, he said. “Because we haven’t looked.”