Songbirds need so-called “human language gene” to learn new tunes

Blogging on Peer-Reviewed ResearchThe nasal screech of Chris Tucker sound worlds apart from the song of a nightingale but both human speech and birdsong actually have a lot in common. Both infants and chicks learn their respective tongues by imitating others. They pick up new material most easily during specific periods of time as they grow up, they need practice to improve and they pick up local dialects. And as infants unite words to form sentences, so do songbirds learn to combine separate riffs into a full song. Songbirds need so-called “human language gene” to learn new tunes

Because of these similarities, songbirds make a good model for inquisitive neuroscientists looking to understand the intricacies of human speech. Zebra finches are a particularly enlightening species and they have just shown Sebastian Haesler that the so-called human ‘language gene’ FOXP2 also controls an songbird’s ability to pick up new material.

FOXP2 has a long and sordid history of fascinating science and shoddy science writing. It has been consistently mislabelled as “the language gene” and after the discovery that the human and chimp versions differed by just two small changes, it was also held responsible for the evolution of human language. Even though these claims are far-fetched (for reasons I’ll delve into later), there is no doubt that faults in FOXP2 can spell disaster for a person’s ability to speak.

Mutated versions cause a speech impairment called developmental verbal dyspraxia (DVD), where people are unable to coordinate the positions of their jaws, lips, tongues and faces, even though their minds and relevant muscles are in reasonable working order. They’re like an orchestra that plays a cacophony despite having a decent conductor and tuned instruments.

Brain scans of people with DVD have revealed abnormalities in the basal ganglia, an group of neurons at the heart of the brain with several connections to other areas. Normal people show strong activation of FOXP2 here and fascinatingly, so do songbirds. Haesler reasoned that studying the role of this gene in birds could tell him more about its human counterpart.

Singing in the brain

When young zebra finches learn to sing, or when adult canaries add to their repertoire in the summer, the activity of FOXP2 in their basal ganglia goes up. It’s particularly active in an ominously named region called Area X that birds need for learning songs, and it is there that Haesler turned his attentions. Basal ganglia

Together with colleagues from the Max-Planck Institute of Molecular Genetics, he injected 23-day old finches with a virus carrying a special piece of RNA designed to be a molecular off-switch. It reduced the production of the FOXP2 protein at Area X by about 40%. After the injection, each bird received private singing lessons from an adult male housed in the same sound-proof cage.

Normally young zebra finches start to memorise the songs of an adult tutor 25 days after hatching. They sound awful at first, but with lots of practice, they do pitch-perfect renditions by the three month mark. In contrast, the infected finches did no better than the average karaoke singer. As they grew up, Haesler analysed their songs and found them to be pale and garbled versions of their tutors’ melodies.

They had the same component syllables but these were rearranged, left out, repeated incorrectly, or sung with the wrong pitch or length. With less FOXP2, the young finches simply couldn’t learn to accurately imitate the song of another finch. As adults, their songs remained garbled, replete with syllables that had the wrong sound or length. They showed remarkably similar problems to children with DVD, and to deaf finches who can hear neither their own songs nor the template tunes of adults.

Clearly, FOXP2 doesn’t control the ability to sing as the impaired youngsters still made the right range of sounds. Nor does it control learning, since the finches learned their tutors’ songs badly, rather than not at all. Their problem lay in being unable to adjust their songs to match those of their tutors. Haesler suggests that the gene affects the development of parts of the brain that allow the finches to learn the right songs by listening, repeating and practicing.

A “language gene” indeed?

This study lands a heavy blow on the already precarious assertion that FOXP2 is a “human language gene” let alone the human language gene. In reality, there’s likely to be a multitude of genes that affect the development of language, and many where mutations cripple our ability to speak intelligibly. It’s certainly speculative to claim that Neanderthals had language because they had the same version of FOXP2 as us, as Greg Laden neatly debunks at his blog.

If the human version works in the same way as the finch one, it doesn’t control the development of language nor the ability to speak. Instead, when children learn to speak, the gene affects the development of the brain so that they can produce the right sounds based on what they hear and learn from their parents.

This theory will need further support, but an earlier study in bats suggests it’s on the right lines. This year, a Swedish group found that the gene is associated with the development of echolocation in bats, where it has diversified far more than in other mammal groups. Since echolocation involves the exquisite coordination of a bat’s squeaks and senses, FOXP2 may once again act as a link between listening and speaking.

Reference: Haesler, S., Rochefort, C., Georgi, B., Licznerski, P., Osten, P., Scharff, C. (2007). Incomplete and Inaccurate Vocal Imitation after Knockdown of FoxP2 in Songbird Basal Ganglia Nucleus Area X. PLoS Biology, 5(12), e321. DOI: 10.1371/journal.pbio.0050321


One Response

  1. […] are often biased and not very entertaining to read. But not in the bird world! Check out Ed Yong’s post on the study of the so-called “human language gene” in bi…. Apparently, this gene, when mutated, causes speech impairment in humans – but the same gene in […]

Comments are closed.

%d bloggers like this: