Margoliash and Yu measured the electric impulses that form the musical instructions by outfitting 13 male finches with special headgear with brain electrodes. The four electrodes were poised above a region called the HVc, known to be part of the bird's song production system, or another region called the RA, a subordinate region that receives input from the HVc and activates the muscles of the syrinx--the bird's voicebox. They recorded electrical activity just outside individual brain cells of interest, and compared the signal with sounds the birds made.

The impulses in the RA coincided with individual notes, each lasting just a few hundredths of a second, while the signature firing pattern from the HVc corresponded to a syllable, or group of notes, lasting in the range of a quarter of a second.

Margoliash said the goal now is to see how the brain learns the song, by comparing adult birds to immature ones or those that are raised in isolation or have diminished hearing.

"What we can hope to find eventually is: How does learning influence the activity patterns of individual neurons? We're learning how the circuits are put together, and we should gain insight into what elements of the circuit are modified during learning."

The research was funded by the Whitehall Foundation, a private philanthropy that funds biomedical research with an emphasis on neurobehavioral studies.