Have you ever been at a noisy party, so engrossed in conversation that background chatter fades and all you hear is the person you’re talking to? Or have you asked your partner a question when the television is on and gotten no response? These experiences demonstrate that our brain can volitionally control what we listen to—a phenomenon known to neuroscientists as “the cocktail party problem.”
How selective attention works remains a puzzle, but recent research from UC Berkeley has provided some clues. Chris Rodgers, a graduate student in Professor Mike DeWeese’s lab, simulated the cocktail party problem using rats. Instead of voices, Rodgers played two synthetic sounds simultaneously and trained the rats to pay attention to one sound while ignoring the other. Instead of cocktails, he rewarded the rats with water when they responded to the correct sound. “The hard part was training me to train the rats effectively,” Rodgers jokes. “Once we knew the right way to train them, they were able to learn the task very quickly.” While they performed the task, Rodgers recorded electrical activity from two regions of the rats’ brains involved in sound processing and decision-making.
“Our main finding was a surprise,” says Rodgers. “We expected to see a difference in the way the different sounds themselves were encoded. Instead, we found a difference in the neural activity before the sound came on: a sort of anticipatory effect.” They speculate that the observed neural activity represents the rat’s memory of which sound it should pay attention to. Interestingly, they found that rats could rapidly switch their attention from one sound to the other. Rodgers equates this to our own experiences, saying “we can switch in a moment from listening to one conversation to eavesdropping on another.”
In addition to understanding the neural basis of eavesdropping, these findings have several potential applications. “There’s a strong medical reason to want to understand this,” says DeWeese. “There are a host of auditory attention disorders, including schizophrenia, autism, and attention deficit disorder.” These findings might also be applied to develop better voice recognition software, capable of separating voices from background noise. And perhaps one day, scientists will discover the best way to train your partner to listen to you when the television is on.