Do re mi fa ... How do you know what comes next?

Brain scientists at Georgetown University Medical Center now have a response to these questions. At Neuroscience 2012, the annual meeting of the Society for Neuroscience, the researchers reveal their solution to what has long been a fundamental puzzle in neuroscience: What does the brain have to do to process a new musical sequence, and what must it do to recall a song, once learned?

The answer, says Brannon Green, a graduate student who works in the laboratory of senior author, neuroscientist Josef Rauschecker, Ph.D., is that two different areas of the brain are used -- one to learn a sequence and another to recall it -- and that higher motor areas participate in both.

Green, Rauschecker, and three colleagues at Aalto University in Helsinki/Finland used functional magnetic resonance imaging (fMRI) with volunteers who listened to musical sequences while in the scanner. The participants heard early parts of a 30-second musical sequence 20 to 30 times, making that part of the melody familiar to them, while they heard the rest of the music only one to 10 times.

The researchers found that learning a novel sequence required use of the brain's motor areas, such as the basal ganglia and cerebellum, which would seem more likely to be used to move muscles to sing. In this study, these areas were active in learning the sequence of the parts of the tune -- that one sound fired one neuron, and the next sound fired a second neuron, and so on.

"The motor system contains brain structures that nature invented to decode sequences, so to learn a melody, the auditory system hijacks the motor system," says Rauschecker.

"This is also the part of the brain you use to learn how to ski or dance," says Green. "This study shows that motor sequencing areas in the brain are really generalized sequencing areas that process both motor and auditory sequences."

Once the study participants learned the tune, the fMRI showed that brain activity switched from the motor regions to areas in auditory and prefrontal cortex, which are associated with long term memory of sounds. And it also showed that it didn't take as many neurons to remember the sequence as to initially learn it.

Rauschecker likens storage of a tune to dominoes stacked next to each other. "The tones are chunked together, chained to one another in a sequence. It's like pushing the first domino, and then they all fall, one after the other. The memory trace is tight; fewer neurons are needed for recall."

"That can explain why if you get stuck in the middle of a melody you are playing on a piano, it is easier to recall it if you start from the beginning of the tune," Green says.

Tipping the dominos at will also illustrates "how a symphony conductor is always ahead of the orchestra," Rauschecker says. "One cue triggers the next and then the next."

The research was funded by the National Science Foundation (PIRE-OISE-0730255) and the Academy of Finland's FiDiPro program. Rauschecker and his co-authors report having no personal financial interests related to the study.