Songbirds stutter, babble when young, become mute if parts of their brains are damaged, learn how to sing from their elders and can even be "bilingual"—in other words, songbirds' vocalizations share a lot of traits with human speech. However, that similarity goes beyond behavior, researchers have found. Even though humans and birds are separated by millions of years of evolution, the genes that give us our ability to learn speech have much in common with those that lend birds their warble.

A four-year long effort involving more than 100 researchers around the world put the power of nine supercomputers into analyzing the genomes of 48 species of birds. The results, published this week in a package of eight articles in Science and 20 papers in other journals, provides the most complete picture of the bird family tree thus far. The project has also uncovered genetic signatures in song-learning bird brains that have surprising similarities to the genetics of speech in humans, a finding that could help scientists study human speech.

The analysis suggests that most modern birds arose in an impressive speciation event, a "big bang" of avian diversification, in the 10 million years immediately following the extinction of dinosaurs. This period is more recent than posited in previous genetic analyses, but it lines up with the fossil record. By delving deeper into the rich data set, research groups identified when birds lost their teeth, investigated the relatively slow evolution of crocodiles and outlined the similarities between birds' and humans' vocal learning ability, among other findings.

The vocal learning discoveries could have important implications for the study of human speech and its disorders. If the genes are similar, "you can study in song birds and test their function in a way you can't do in humans," says Erich Jarvis, one of the leaders of the international effort and an associate professor of neurobiology at Duke University.

Scientists have long used songbirds, typically zebra finches, to probe questions about how language can be learned because not many other species have this ability. "Among primates, Homo sapiens are the only species that can modify vocalization," says Stephanie White, a neuroscientist and professor of integrative biology and physiology at the University of California, Los Angeles, who was not involved in the new research.

That's not to say that other primates don't communicate vocally, but White explains that the grunts, screeches and hoots uttered by chimpanzees, for example, are more automatic. Although an older, bigger chimp may have a deeper voice, "a young chimp and an old chimp sound pretty much the same," she says. Humans and songbirds, on the other hand, progress from baby talk to complex vocalizations. The handful of other species with this ability—the vocal learners—includes dolphins, sea lions, bats and elephants.

The new work on vocal learning relied on laser dissection of brain regions of zebra finches known to be involved in vocalizations and then analysis of gene activity there. The researchers then compared those levels to gene expression levels in human brains. They found that humans and birds share 55 genes  between brain regions important for vocal learning, a good handful of which were involved in forming connections between neurons. Analysis of genes in other avian vocal learners — parrots and hummingbirds—echoed the finding.

Another paper shows that 10 percent of the genome in song-learning birds is dedicated to song. White, who found both papers to be "very powerful," explains that these genes are actively regulated during vocalization. In humans, a simple phone conversation is actually an intensely focused activity that sets off cascades gene regulation across the brain, she says.

The similarity of the gene networks needed for vocal learning between humans and birds is not completely surprising. After all, all vocal-learning species can trace their ancestry back to the same basal branches on the tree of life, White says. Even though the ability evolved independently, it was influenced by a similar initial deal from the genetic deck of cards. Also, the broadly similar environment of this Earth created the evolutionary pressures that shape vocal learners. Just as multiple species came up with similar solutions to the problem of vision, species that evolved vocal learning seem to have settled on common strategies.

Viewed from another angle, however, the convergence is striking. "This, to my knowledge, is the first time a learned behavior has been shown to have so much similar molecular underpinnings," White says. The discoveries open up a host of potential avenues for future exploration: Can nonvocal learners acquire some traits needed for vocal learning simply by tweaking some key genes? Almost certainly, zebra finches have more to tell us about our own ability to babble, shout and sing.