Humans Are Birdbrained When Learning Speech, Study Hints

John Roach
for National Geographic News
Updated May 17, 2006
Hummingbirds are well known for their ability to flap their wings at an
eye-blurring 75 beats or more per second. Less known, perhaps, is the
fact that they can learn to sing the hummingbird equivalent of

Like parrots and sparrows, whales and dolphins, and bats and humans, hummingbirds are part of a select group of animals that possess the ability to imitate and learn sounds—a process known as vocal learning.

Erich Jarvis, a neurobiologist at Duke University in Durham, North Carolina, studies the vocal-learning ability of birds like hummingbirds and zebra finches. His aim: to understand how the brain pathways for vocal learning evolved and work.

Most recently, Jarvis collaborated with two international teams of scientists who share this research interest. The scientists found that vocal-learning birds carry a gene called FoxP2, which becomes active when the birds are learning a new song.

The researchers note that a mutation in a nearly identical gene in humans causes an inherited language deficit.

The findings led the scientists to conclude that FoxP2 "is an important component to the circuitry of vocal learning," Jarvis said. "We can't say it is required, but we can say it is being used."

The teams described their findings in two papers published in the Journal of Neuroscience in 2004.

Jarvis and his lab colleagues led one of the studies with co-author Constance Scharff of the Max Planck Institute for Molecular Genetics in Germany. Stephanie White, a professor of psychological science at the University of California, Los Angeles, led the other study.

In future work, Jarvis said, he hopes to breed songbirds without FoxP2 to determine whether or not the gene is required for vocal learning. That finding may have implications for better understanding human speech and language.

Marc Schmidt is a neurobiologist at the University of Pennsylvania in Philadelphia. He said Jarvis's work pushes scientists to see parallels in the underlying organization of neural circuits involved in vocal learning between songbirds and humans. The two animals sit on two distinct branches of the evolutionary tree.

Schmidt said, "[Jarvis] is bold enough to cross disciplines and make people realize that these systems are organized in very similar ways."

Independent Evolution

Jarvis says scientists seeking to understand human vocal learning would traditionally study the process in an animal more closely related to people, like a chimpanzee. Jarvis notes, however, that nonhuman primates lack the ability to imitate and learn sounds.

Scientists believe that vocal learning evolved independently in three kinds of birds (songbirds, parrots, and hummingbirds), humans, bats, and marine mammals (like whales, dolphins, and porpoises).

Since the trait evolved independently, some scientists have argued it is not possible to compare vocal learning in one animal to vocal learning in another.

Jarvis says modern scientific techniques like brain scans and gene cloning, however, suggest the pathways for vocal learning in bird brains may be similar to those in human brains.

A good analogy, Jarvis notes, is wings, which evolved independently on birds and bats. On both creatures, wings are located in similar body locations and share many characteristics.

"It suggests there are physical constraints on the organisms' interaction with the environment," Jarvis said. "If you are going to evolve wings, nature says it is done this way. Same thing with vocal learning."

Vocal Learning

Dogs bark, cats meow, and bears growl. The ability for these animals to make each of their signature sounds is hardwired from birth. But scientists say none of these animals can learn to bark, meow, or growl in a new way. They are not vocal learners.

By contrast, songbirds, humans, whales, dolphins, and bats are born with the ability to make certain sounds. As individuals age and mature, they learn how to modify those sounds to make new sounds, Jarvis said.

"Humans are born with an innate set of phonemes"—the basic building blocks of phonetic sound used in language, like the g in "goal"—"which we then modify to strings of phonemes, to make words, and strings of words, to make sentences," Jarvis said.

Like humans, songbirds modify calls (the birds' equivalents of phonemes, or bits of sound) and put them into a sequence to make songs. Cetaceans—whales, dolphins, and porpoises—are thought to do the same with their innate sounds.

According to Jarvis, the process of vocal learning is one of passive imitation. Children listen to adults and with partial success imitate what they hear. As children mature, they go through a stage of "crystallization," in which their imitations become accurate.

Vocal learning for humans gets harder after puberty, which is why educators say it is best to learn a second language while in grade school. Yet humans never completely lose the ability to learn new vocalizations, Jarvis says.

Of the more than 4,000 different songbirds, some, like canaries, can learn new songs their entire lives. Others, like zebra finches, find it impossible to imitate a new song after their puberty-like phase in life. These species are called closed-ended vocal learners, said Jarvis.

By studying the differences among songbirds, hummingbirds, and parrots, Jarvis and his colleagues gain more insight into the process of vocal learning and ultimately the process of how the brain, in general, generates behavior.


In 2002 a research team led by Anthony Monaco at the University of Oxford, England, found that the FoxP2 gene is associated with language production in humans. That finding led Jarvis and colleagues to look for the gene in other vocal-learning animals.

The researchers found the FoxP2 gene in a range of vocal-learning birds, including finches, canaries, and hummingbirds. They also found it in the nonvocal-learning ring dove and crocodile, the closest living relative to birds.

The researchers then compared the FoxP2 gene in songbirds to the FoxP2 gene in humans. The scientists found that two isolated mutations in the human gene were not present in the birds, suggesting that the mutations are not required for vocal learning.

"The third question we asked—mutation or not—[was], Is there something about the FoxP2 gene in vocal learners that is unique to vocal learners?" Jarvis said. "The answer is yes. There are unique patterns of expression in FoxP2 in the brains of vocal learners."

Jarvis and colleagues found that levels of the gene expression increase right before animals learn to imitate new sounds. The gene expression, meanwhile, falls off when animals are not learning. This suggests that FoxP2 is associated with a behavioral plasticity that makes vocal learning possible.

Schmidt, the University of Pennsylvania neurobiologist, said the paper pushes the idea that "neural circuits, whether in humans or birds, may need to adopt common strategies in order to become functionally capable for vocal learning."

Free Email News Updates
Sign up for our Inside National Geographic newsletter. Every two weeks we'll send you our top stories and pictures (see sample).


© 1996-2008 National Geographic Society. All rights reserved.