Skip navigation
Songbird’s strategy to change its tune could inform rehab efforts


For over half a century, neuroscientists have been gaining insight into the human brain by studying songbirds. That's because like humans, songbirds are able to learn very complex vocalization.

We're interested both from a basic point of view of understanding how it is that we do this very complex thing that's very important for all aspects of our life and we're also ultimately interested in understanding learning from a perspective of being able to create tools that enable rehabilitation or repair in cases where learning has essentially gone awry or where there's been damage to the brain that prevents normal functioning.

Jonathan Charlesworth works in Michael Brainard's lab at the University of California, San Francisco. He has been using a simple, automated computer system to figure out how songbirds learn to perfect and maintain their song and hopes to gain better insight into how the brain changes during the learning process.

So, the general goal of this research was to improve our understanding of how the nervous system learns from signals of success or failure.

In his experiment, the computer program emitted an unpleasant sound while songbirds sang, which caused them to learn to change their tune. They did so by remembering every slight change in pitch of a single syllable sung about 500 times a day.

The general result here means that the brain keeps track of behavior in a more detailed way than we thought before. If you're trying to teach somebody something, a fine motor skill like playing the piano or speech, or you're trying to re-teach them something, like they originally knew how to do it but now they've become unable to through some sort of  injury or illness. You don't have to give feedback signals that are extremely complicated. You don't have to tell them exactly what they're doing wrong. You just have to tell them whether this was good or bad and the brain can do most of the rest of the work.

Our results suggest that if the principle we've studied in the songbird apply to humans, learning might be greatly facilitated by taking advantage of this very narrow time window after we try out a vocal gesture to provide feedback about whether it's been successful or not. And we think that by studying some of these very general principles in a simpler system, we can gain knowledge that's quite relevant to understanding much more broadly how learning works and how rehabilitation might be best carried out.