Singing and the Brain

Humans have a special ability to learn songs and change them over time—a skill that is useful in treating neurological problems and may help support overall brain health. Researcher Aniruddh D. Patel explains.

Birds do it. Bees do it. Even itty-bitty fruit flies do it. They sing, and often quite beautifully. Just think of the intricate and mournful song of a loon heard on the lake at sundown.

Human song connects us, not just to each other, but to other species. But there are key differences in the way humans and other animals sing, and those differences point to the unusually important role that sound plays in human brain function.

Aniruddh D. Patel, a researcher at Tufts University and self-described “neuroscientist of music,” studies how music and singing affect the human brain. During his popular session at Chorus America’s 2015 Conference, he shared the latest research on how singing may benefit brain function and the relationship between music and language.

What Makes Human Song Special?

For most creatures, their song changes very little over time. “Most animal songs are innate, which means they are born with them. It is their signature sound,” says Patel. Except for a few species of birds, such as parrots and hummingbirds, and certain sea mammals, such as beluga whales, the ability to learn songs is uniquely human.


Dr. Aniruddh D. Patel

The fact that humans have the capacity to learn songs at all is “bizarre” from an evolutionary perspective, Patel says. Humans are primates, with the same forward-facing eyes and great depth and color perception that helps all primates navigate their environments. “Compared to most mammals, primates are visual, not auditory, specialists,” Patel says. But the human brain differs from the brains of other primates because of the prominence of our temporal lobes—areas of the brain around the temples that are involved in auditory processing.

Our heightened ability to analyze sounds may have been adaptive. Just as parrots use their learned songs as a way to identify themselves as belonging to a specific flock, “human vocal learning may have started as a way to mark ourselves as being members of a group, maybe before we had full-blown language,” Patel says. He connects this group identification with the way accents tell us what region or country a speaker comes from today. “One of the interesting things about accents is you learn them early and they are very hard to change. They are a strong identifier of early community.”

Singing and Patients with Neurological Issues

The fact that sound is such a powerful part of human brain functioning has opened up rich areas of research. One area is the use of song to improve brain function for people with neurological issues.

Nonfluent Aphasia. People who have suffered damage in their brain’s left hemisphere after stroke often have problems producing language—a condition called nonfluent aphasia. “Patients may have trouble putting a three-word sentence together and yet if you ask them to sing an old song, they will sing it fluently,” Patel says. “You wouldn’t know there was anything wrong with them.”

The phenomenon, described memorably in Oliver Sacks’ book, Musicophilia, suggests that there are specialized areas of song in the brain that are not damaged in these patients. Comparing Magnetic Resonance Imaging scans (MRIs) bears this out. “There are distinctive areas, more on the right side of the brain, that are involved with singing and you can see these ‘light up’ in brain scans of healthy people when they are producing or perceiving song” Patel says. “There is a right hemisphere bias in the control of song.”

“Music provides a way to access regions of the brain and reawaken autobiographical memory when language won’t.”  -Aniruddh D. Patel

In the 1970s, a new therapy sought to harness the power of song to help patients with nonfluent aphasia improve their speech abilities. Melodic Intonation Therapy (MIT) creates short, simple melodies with a rhythmic component that a patient can use to intone basic phrases, such as “I love you” or “I love my children.” The therapist models the melodies and the patient sings along and taps out the rhythm. “The phrases gradually get longer and the goal is ultimately to get the patient to say what they want to say, not just stock phrases,” Patel says.

In an ongoing study at Harvard Medical School led by Gottfried Schlaug, MIT is being compared to speech repetition therapy, a conventional treatment in which patients repeat phrases to the therapist without the melodic component. Preliminary results have shown that both groups improved in speech ability, but in the MIT group there was an improvement of 200 percent, compared to an improvement of 160 percent in the speech repetition therapy group. “The conventional wisdom is that more than a year after a stroke, patients have recovered as much as they will,” Patel says. “But with both therapies there were additional improvements—much more so with the melodic therapy.”

Brain imaging confirmed what the researchers were observing. “It looked like right the hemisphere brain regions involved in singing had taken over for some of the damaged parts in the left hemisphere,” Patel says, “and had been recruited for speech.”

The researchers also looked at connections between different regions in the brain to see if those connections had been changed or strengthened by the Melodic Intonation Therapy. They focused on one long-distance connection from the front to the back of the brain that is important for speech and song called the arcuate fasciculus. “They found some evidence that melodic therapy strengthened that connection in the right hemisphere,” Patel says. “There were some changes in the size of this brain pathway on the right side of the brain, suggesting that enhanced auditory-motor communication in the right hemisphere may help compensate for damage to this pathway in the left hemisphere.”

Dementia. Research has also found evidence that musical memory is retained, even in severe cases of dementia and Alzheimer’s disease.

“Patients might not know your name,” says Patel, “but if you play familiar songs to them, they might sing along, and know the melody in great detail. If you introduce a wrong note into the song, they will startle, which is evidence they know the details of that melody.”

This can be important for people with dementia, Patel says, because “musical memories have a strong link to autobiographical memories. The music we remember from our youth often reawakens memories of where we were in our lives, who we were with, what we were like, what we were going through.”

But why is musical memory preserved in dementia? Again, studies using brain imaging can help explain. Studies by Petr Janata and colleagues at UC Davis found that music that had the qualities of being familiar, pleasing, and evocative of autobiographical memories was processed predominantly in the dorsomedial prefrontal cortex. This is among the areas of the brain that are slower to atrophy with Alzheimer’s disease.

“Could singing be neuroprotective?” he asks. “If you’re a singer and have a stroke, would you recover faster? Could regular singing have a lasting impact on brain function? It’s an idea that should be studied more.” -Aniruddh D. Patel

“Music provides a way to access regions of the brain and reawaken autobiographical memory when language won’t,” Patel says.

Early results suggest that the strongest effects of music therapy may be on patients’ emotions, particularly their anxiety levels. “The music does not ameliorate the cognitive effects of the dementia,” Patel says, “but it does have a strong impact on a patient’s emotional state, and potentially on quality of life.”

Can Singing Also Improve Brain Function in Healthy People?

The research about the effects of song on people with brain conditions raises hope for more far-reaching impact. “If this works for patients, could it work for ordinary people to help with their brain health?” Patel asks. To address that question, Schlaug and his team have done MRIs of older healthy adults, dividing them into a group of singers and a group of non-singers.

The group of singers showed greater connections between areas of the brain than the non-musician group, with the strongest difference on the left side. “Song combines music and words, and word production is a left hemisphere-biased activity,” says Patel. “Perhaps doing lots of singing strengthens the brain networks involved in word production and articulation, in addition to the right hemisphere circuits involved in fine control of pitch and melody.”

These findings suggest new research possibilities, Patel says. “Could singing be neuroprotective?” he asks. “If you’re a singer and have a stroke, would you recover faster? Could regular singing have a lasting impact on brain function? It’s an idea that should be studied more.”

Using Prediction to Study Language-Music Relationships in the Brain

Another research area that Patel is heading up at his lab at Tufts University is studying the relationship between singing and prediction—the ability to anticipate what comes next. A longtime topic of research in music cognition, prediction has also recently become a growing focus of research in language processing. That makes it a good topic for investigating language-music relationships in the brain, a major theme of Patel’s research.

As Leonard Meyer wrote in his 1956 book, Emotion and Meaning in Music, the “fulfilling or thwarting” of expectations is a principal source of music’s emotional power. In his 2006 book Sweet Anticipation: Music and the Psychology of Expectation, David Huron shows how common musical devices (such as syncopation, cadence, meter, tonality, and climax) exploit basic psychological mechanisms of expectation. For example, successfully anticipating the outcome of a particular musical phrase can trigger positive feelings—even more so if the gratifying outcome is delayed by a slower tempo or another compositional technique.

The OPERA Hypothesis
Aniruddh Patel’s OPERA hypothesis proposes that singing and musical training benefit language skills because musical training helps develop speech-processing networks in the brain when the following five conditions are met:
  • Overlap in the brain networks that process an acoustic feature used in both music and speech
  • Precision: music places higher demands on these shared networks than does speech, in terms of the precision of processing
  • Emotion: the musical activities that engage this network elicit strong positive emotion
  • Repetition: the musical activities that engage this network are frequently repeated
  • Attention: the musical activities that engage this network are associated with focused attention.

The ability to predict what comes next is also central to the way we experience language. “When you are listening to people talk, you don’t just passively soak in information,” Patel says. “You are implicitly predicting the words that people are going to say next. That can help in understanding people in noisy environments or where things are slightly ambiguous.”

While expectation doesn’t seem to have the same strong link to emotion in language processing, both music and language involve complex sequences that unfold rapidly in time in which elements vary in predictability. Patel and his team wanted to know: do music and language share mechanisms for prediction? “The old chestnut is that music and language are quite different brain functions,” Patel says. “The left brain is more verbal and logical, and the right brain is more involved in art, music, emotion. But in fact, there is a lot of research that indicates there is much more overlap than we had thought.”

To explore the possible overlaps, Patel and his team are conducting an experiment where subjects hear pairs of melodies and pairs of sentences, some of which lead to strong predictions and some that don’t. An example of a sentence pair is:

1) “When the two met, one of them held out his _________”

2) “When the two met, one of them brought his _________ ”

The first sentence leads to a strong prediction: most people use the word “hand” to complete it. The second does not, as people vary in the word they use to complete the thought.

An example of a melody pair is:

Melody A

Melody B

The first melody leads to a strong prediction. Since it ends with an implied authentic cadence, most people sing the tonic (D) as a continuation of this melody. Although the second melody is identical to the first in length, rhythm, melodic contour, musical key and number of notes, people vary widely in the note that they sing as a continuation. The second melody doesn’t lead to a strong prediction.

The idea is to compare patterns of brain activity as research subjects process predictions for sentences versus melodies to see how these patterns are similar or different. During the experiment, electrical activity in the brain is measured by using sensors attached to the subject’s heads.

Can Musical Training Improve Language Skills?

In another study, Patel plans to examine whether musical training enhances linguistic prediction abilities. “If you are musically trained, are you better at predicting what’s coming up next in a sentence?” he asks.

This could have implications for language skills for young and old. “For older adults in crowded environments, where understanding speech is challenging, better prediction abilities might help them anticipate which word is coming next,” Patel says. “If children who are learning to read are good predictors of upcoming information in sentences they may be able to more quickly integrate the words they encounter and become more efficient readers.”

These studies of prediction are part of Patel’s broader theory on how musical training impacts speech processing, called the OPERA hypothesis. The OPERA hypothesis suggests that because music and speech share some cognitive processing mechanisms in the brain, and music places

This article is drawn from Aniruddh D. Patel’s presentation at Chorus America’s 2015 Conference in Boston and a follow-up interview. Patel joined Tufts University in the fall of 2012 as an associate professor of psychology. Previously he was a senior fellow at The Neurosciences Institute in San Diego. He is the author of Music, Language and the Brain (Oxford University Press), which won a Deems-Taylor award from ASCAP in 2008 and the 2009 recipient of theMusic Has Power Award from the Institute for Music and Neurologic Function in New York City.
Kelsey Menehan is a writer, psychotherapist, and longtime choral singer based in San Francisco.

higher demands on these mechanisms than speech does, musical training may enhance language skills.

As Patel’s research shows, singing is giving us new tools for comparing ourselves with other species, for neurological rehabilitation, and for understanding how we interact with our world. “The study of music and the brain is exciting because it addresses very ancient questions about the power of music using cutting-edge methods,” he says. “The field has potential for laying the foundation for therapies that enhance brain function in people with neurological disorders, and for illuminating basic aspects of human brain function.”


Want to Learn More?

The Great Courses, an adult learning resources, recently released Aniruddh Patel’s “Music and the Brain” course. This series of 18 half-hour lectures covers fundamental ideas of music theory, neuroanatomy, and cognitive science and looks at the diverse range of experiments, discoveries, and debates in this fast-changing field. Access the program online at thegreatcourses.com.


Resources

 

Books
Emotion and Meaning in Music by Leonard Meyer, University of Chicago Press, 1956. http://press.uchicago.edu/ucp/books/book/chicago/E/bo3643659.html

Music, Language and the Brain by Aniruddh Patel. Oxford University Press, 2008. http://www.amazon.com/Music-Language-Brain-Aniruddh-Patel/dp/0199755302

Musicophilia: Tales of Music and the Brain by Oliver Sacks, Random House, 2007.
http://www.amazon.com/Musicophilia-Tales-Music-Revised-Expanded/dp/1400033535/ref=sr_1_1?s=books&ie=UTF8&qid=1437766870&sr=1-1&keywords=musicophilia

Sweet Anticipation: Music and the Psychology of Expectation by David Huron, MIT Press, 2006. https://mitpress.mit.edu/books/sweet-anticipation

Articles
“Melodic Intonation Therapy: Shared Insights on How it is Done and Why it Might Help” by Andrea Norton, Lauryn Zipse, Sarah Marchina, and Gottfried Schlaug, Annals of the New York Academy of Science, July 2009. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2780359/

“The Neural Architecture of Music-Evoked Autobiographical Memories” by Petr Janata, Cerebral Cortex, Nov. 19, 2009. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2758676/

“Why musical memory can be preserved in advanced alzheimer’s disease” by Jörn-Henrik Jacobsen, Johannes Stelzer, Thomas Hans Fritz, Gael Chételat, Renaud La Joie, Robert Turner, Brain, June 3, 2015. http://brain.oxfordjournals.org/content/early/2015/06/03/brain.awv135

“Why would musical training benefit the neural encoding of speech? The OPERA hypothesis” by Aniruddh Patel, Frontiers in Psychology, June 29, 2011.

Documentaries
The Music Instinct: Science and Song by Elena Mannes
https://www.youtube.com/watch?v=wZmMF6cprC0

Organizations
Society for Music Perception and Cognition
www.musicperception.org