Scientists have long struggled to understand how human language evolved. Words and sentences don’t leave fossils behind for paleontologists to dig up.
A genetic study published on Tuesday offers an important new clue. Researchers found that, between 250,000 and 500,000 years ago, a gene known as NOVA1 underwent a profound evolutionary change in our ancestors. When the scientists put the human version of NOVA1 into mice, the animals made more complex sounds.
Erich Jarvis, a neuroscientist at Rockefeller University and a co-author of the new study, cautioned that NOVA1 alone did not suddenly switch on our ancestors’ language abilities.
“I wouldn’t say it’s ‘the’ language gene,” Dr. Jarvis said.
Instead, over millions of years, language arose thanks to mutations in hundreds of genes.
“But where does NOVA1 fit into that whole combination? It’s one of the last steps,” Dr. Jarvis said.
NOVA1 drew scientific attention in 2012 when it appeared on a special list of genes that produced proteins that were identical in most mammals, but produced a different form in humans. Out of more than 20,000 protein-coding genes, only 23 made the list. All were probably crucial to the evolution of our species.
That NOVA1 made the list surprised Dr. Robert Darnell, a neuroscientist at Rockefeller University who had discovered the gene in 1993. He was startled because the gene appears to be essential to all mammals. A mouse engineered without NOVA1 will die during development. Nothing in Dr. Darnell’s research had hinted that the gene had played a distinctive role in human evolution.
Dr. Darnell began collaborating with evolutionary biologists to find out more. One of them, Adam Siepel of Cold Spring Harbor Laboratories in New York, led an effort to reconstruct the gene’s history. He looked at the gene’s sequence in the DNA of extinct humans, as well as genetic information from more than 650,000 living people.
Dr. Siepel found that NOVA1 underwent a dramatic change not long after our ancestors split from Neanderthals and Denisovans, providing an evolutionary advantage to early humans who inherited the change. Eventually, it swamped the original version of NOVA1.
Ever since, Dr. Siepel said, that version of the gene has remained overwhelmingly dominant in the human population. Mutations that reversed NOVA1 to its original form must have been harmful, because they are extremely rare. Of the 650,000 people registered in the database, only six carried the original version of the gene.
The researchers do not know anything about who those six people are. Dr. Darnell is now searching for carriers of the original NOVA1, in the hopes of testing them for speech skills.
In the meantime, Dr. Darnell and his colleagues have engineered mice that carry the human version of NOVA1, instead of the one found in other mammals. To all outward appearances, the NOVA1 mice seemed ordinary. But they harbored some telling differences.
The human version of NOVA1 oversaw the production of 200 proteins in mouse brains that the ordinary version of the gene did not. And many of those proteins played a role in how the animals produce sounds.
“For me, that was like, ‘Bingo!’” Dr. Darnell said.
If NOVA1 had shaped the evolution of human language, Dr. Darnell reasoned, then the human version of it might change the way that the mice produce sounds. Dr. Jarvis, an expert on animal vocalization, helped Dr. Darnell eavesdrop on the animals.
Mice typically produce pulses of ultrasonic squeaks that resemble syllables in human language.
But mice carrying the human version of NOVA1 made peculiar squeaks, the scientists found. The difference was especially noticeable when males sang courtship songs to females. Their songs contained more complex sounds, and the mice switched between those sounds in more intricate patterns.
The intriguing changes in NOVA1’s evolution happened after our ancestors split from Neanderthals and Denisovans. But another language gene, known as FOXP2, underwent a burst of important changes before that split. And studies have shown that mice carrying human FOXP2 genes also make strange squeaks.
Some scientists speculated that the two genes both independently altered human brain regions that produce complex sounds.
“The exciting thing about NOVA1 is that there is now another kid on the block,” said Wolfgang Enard, a geneticist at Ludwig Maximilian University of Munich who worked on the FOXP2 mice.
Dr. Jarvis said that he thinks the common ancestor of modern humans, Neanderthals and Denisovans could talk, perhaps thanks to genes such as FOXP2. But mutations to other genes, including NOVA1, may have endowed modern humans alone with the ability to produce a wider range of complex sounds, expanding the power of language.
To test that hypothesis, Dr. Jarvis hopes to engineer mice with mutations in NOVA1, FOXP2 and other genes that may have been important in the rise of language. Together, these mutations may let mice produce even more complex calls.