Listening to the songs of birds

We readily understand "Hello, how are you?" whether the question is posed in a small child's squeaky soprano or large man's booming bass. One way our brain enables this feat is by grouping continuous series of sounds into discrete categories, such as the syllables of a conversation. Despite the central importance of this perceptual process to vocal communication, the underlying brain mechanisms remain largely unknown. Researchers at Duke University Medical Center, recording brain activity in wild swamp sparrows listening to the songs of other birds, now have identified neurons likely to underlie categorical perception of vocalizations.

"The sparrow relies on categorical perception to recognize the songs of other sparrows and thus presents an excellent opportunity to understand how the brain enables categorical perception of vocalizations," says Richard Mooney of the Duke Department of Neurobiology, senior author of the study in Nature Neuroscience would be published online Jan. 11.

Previous studies showed that sparrows use categorical perception to rapidly recognize other sparrows' songs, reacting differently to a song as when the duration of a certain note in that song exceeds a "magical" perceptual boundary.

Using a miniature recording device, Dr. Jon Prather, a postdoctoral fellow in Mooney's lab and the first author of the study, recorded neural activity in a region of the sparrow's brain important to singing, and played subtly different songs to the bird through a nearby speaker. The results were surprisingly clear. The neurons responded briskly as long as the duration of a single note within the song was below a certain length, but not at all when the duration was increased beyond this boundary by even a few milliseconds.

There was an unexpected twist, however. "The neural response boundary was in the wrong place- it simply didn't line up with the reported perceptual boundary," Mooney said. "Interestingly, our birds were from Pennsylvania, and the previous study used birds from New York, and these birds learn to sing slightly different song dialects." This difference raised the possibility that the perceptual boundaries also differed between the two populations. Indeed, by conducting a parallel field study, the team determined that the perceptual boundary in Pennsylvania sparrows differed from that of the New Yorkers and accurately matched the neural response boundary. "This study provides the first neural correlate of vocal perceptual categories and also strongly suggests that learning plays an important role in shaping neural properties for this form of perception," Mooney says.

That such differences in brain and perception could arise in different regions may come as little surprise to natives of Pennsylvania and New York, but is important because it says that both features are affected by learning. In that regard, sparrows and other songbirds are particularly attractive for studying neural processes underlying learned forms of vocal communication, such as speech, because songbirds are one of the few non-human animals that learn to vocalize.

Posted by: Kelly