“Nothing shows that Neandertals didn’t have language abilities,” says Johannes Krause of the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. Indeed, the recent finding by Krause and his colleagues that Neandertals and humans have the same version of the gene FOXP2—the only gene linked to language so far—might be thought of as evidence that they did.

But although studies of modern humans suggest that FOXP2 is necessary for speech, no one believes that it is sufficient. The gene is “just one piece of a complicated puzzle,” says geneticist Simon Fisher of the University of Oxford, part of the team that discovered it. As such, the Neandertal sequence is interesting but provides little information about their linguistic skills. “No single genetic factor can tell us whether or not an extinct species was capable of speech,” Fisher says.

Despite several years of intense study, researchers are still unsure what FOXP2 does or how it might have contributed to the evolution of language. Studies of extinct humans have yet to uncover much about these things, but studies of living animals are starting to provide some hints. The gene has been implicated in many of the most sophisticated sounds in the animal kingdom, suggesting that these behaviors are the foundation on which human language is built.

An English family known as “KE” revealed the link between FOXP2 and language. Many members of this family have severe difficulties with language. They struggle to control their facial movements and have difficulty with reading, writing, grammar and understanding others.

In 2001 Fisher and his colleagues found that the gene at the root of the family’s trouble is FOXP2, located on chromosome 7. The gene makes a protein that binds to DNA, switching other genes on or off. Last month Fisher’s group published a study identifying the 100 genes whose activity is most strongly influenced by FOXP2. Many of the genes turn out to be involved in the development and organization of the nervous system.

But most speech disorders—which affect about 5 percent of all children and have a strong hereditary component—do not involve mutations in FOXP2. They more commonly involve interactions between many genes and environmental factors, says Barbara Lewis, who studies communication disorders at Case Western Reserve University. “It’s a very important gene, but it’s not the only speech gene out there,” she says.

The idea that changes in FOXP2 might have driven the evolution of language got a boost from the finding that the chimpanzee and human versions of the protein differ by two amino acids. This disparity might not sound like much, but FOXP2 is one of the least variable vertebrate proteins. There is only one amino acid difference between the mouse and chimpanzee forms, which diverged 60 million years ago, compared with the six-million-year date for the human-chimp split.

On the other hand, more recent evidence has muddied the link between FOXP2 and language evolution. For instance, the mutations that cause speech defects in humans do not affect those parts of FOXP2 unique to us. And “some of the changes in humans previously thought unique are seen in other mammals,” such as cats, says Stephen Rossiter of Queen Mary, University of London. “As we’re looking at more species, we’re seeing more differences. The picture is getting complicated.”

In September, Rossiter and his colleagues revealed that bats, which use echolocation are an exception to the rule of FOXP2’s unchanging nature: the gene varies widely within the group. “There’s double the number of changes within bats as compared with all the other vertebrates surveyed,” he says. The finding supports the idea that human FOXP2 is particularly important in the physical control of speech. Like talking, which engages more than 100 muscles, making the sounds needed for sonar requires “massively complex coordination of the face and mouth,” Rossiter says.

Bats are one of the few animals that show vocal learning, along with humans, some songbirds, whales and dolphins. That is, the sounds they make are not innate but require practice and imitation. Studies in songbirds support the link between FOXP2 and vocal learning, suggesting that as well as controlling how our brain forms, the gene might also influence how we use it. The gene changes its activity in the brains of adult birds when they learn and practice their songs, neuroscientist Stephanie White of the University of California, Los Angeles, has found. “Birds may have the same circuitry that formed the foundations for human language,” White explains. The evidence points to FOXP2 being a switch, she says, that different species put to varying uses in their neural machinery.

The gene’s story hints at how evolution puts old materials to new uses, points out psychologist Gary Marcus of New York University. “It’s a very good entrée into language and how it relates to whatever preadaptations for language we inherited from our ancestors.”

Doing the FOXP2 Trot
Johannes Krause was surprised when Neandertals turned out to have the same version of FOXP2 as humans. His previous studies of the genetic variation in modern populations had suggested that the human form of the gene arose within the past 200,000 years—150,000 years after Neandertals and modern human lineages diverged. It shows, he says, that our understanding of how current genetics reflects our evolutionary past may need revising.

But Krause still thinks that FOXP2 has been under recent selection pressure in the human lineage. It may have changed just before Neandertals and humans split, he speculates, or it might have helped make our common ancestor more intelligent. “In the fossil record 500,000 years ago, there’s a huge increase in brain size,” Krause notes. “It’s hard to say if it was anything to do with FOXP2, but something was going on.”