They’re playing our song

Elizabeth Margulis looking through a window.

They’re playing our song

Elizabeth Margulis builds bridges between music, imagination and memory

By Alaina O’Regan

What do you think about when you hear your favorite song? Maybe it stirs up vivid memories of an old romance, or you imagine the flames and scent of a bonfire on a humid summer evening.

It turns out that the images that we see in our minds while listening to music are not as individualistic as we might think. People from similar cultural backgrounds tend to visualize similar stories while listening to music, according to a study published in January 2022 in the Proceedings of the National Academy of Sciences by Elizabeth Margulis, professor of music at Princeton, and international colleagues.

“There’s one track we play where people tell us, again and again, that it sounds like there’s a man in a city in the 1920s at night who spots a woman down the street and romantically pursues her,” Margulis said. “There’s another track where most people imagine little animals waking up and starting to frolic.”

Margulis and her group in Princeton’s Music Cognition Lab use music as a tool to study human creativity and memory, and their findings are helping to explain the similarities and differences in what we daydream about while we listen to music. Their studies of how music connects us to our past could also be key to understanding how memories can be preserved into late stages of dementia.

“Getting people to talk about their subjective, internal experiences as they listen to music can help us scientifically understand the relationship between what’s happening with the sound and what’s happening in their brains,” Margulis said.

A natural connection

Combining research from the humanities and sciences, Margulis teamed up with Uri Hasson, professor of psychology and neuroscience at Princeton, to investigate which neural mechanisms we rely on to imagine vivid scenes while listening to instrumental music.

By exploring the neuroscience of human communication, Hasson has pioneered new methods of computing that make it possible to analyze people’s responses to complicated scenarios, like a detailed story or movie. His new methods make it possible to move beyond the typical “beeps” and “boops” played during traditional brain scan studies.

“You can think of music as a modality of human communication,” Hasson said. “For this new project, we want to ask, does the same mechanism by which we communicate using language also govern the ways we communicate using music?”

Understanding how music cues imagination and memory could open paths to treatment for Alzheimer’s disease and other forms of dementia.

To find out, Margulis and Hasson played excerpts of music that they already knew would make people imagine certain stories —like the one that inspired thoughts of animals frolicking — to a group of volunteers undergoing brain scanning with functional magnetic resonance imaging (fMRI) machines, which detect blood flow to various parts of the brain, indicating increased activity. The team then asked the volunteers to tell them what they imagined as the music was played.

Next, the researchers brought in a new group of volunteers and played for them an audio narration drawn from the same words that the people in the previous group had used when describing what they had imagined. By comparing what parts of the brain light up for people daydreaming under different circumstances, the team aims to find out exactly which brain mechanisms are used to imagine stories while listening to music and while listening to narration.

The science of creativity

Although most people from similar cultural backgrounds imagine similar stories, some people consistently come up with something different. Why do some people imagine more conventional stories, while others come up with more outlandish, unusual stories?

To better understand human creativity, Margulis is exploring the root of cultural and individual differences in imagination.

“People often think of imagination as scientifically intractable, like it’s this unbounded creative capacity,” Margulis said. “But because we’ve seen that music guides imagination in a particular way, we see that imagination is constructed out of past experience, memory and perception, and has these very predictable components that we can study and understand.”

To disentangle cultural familiarity from individual imagination, Margulis and her collaborators played music for people in two geographic locations in a musical scale that would be unfamiliar to each, the Bohlen-Pierce scale. It uses an alternative tuning pattern and has different rules about how notes are played together.

The researchers looked at how people’s individual differences, like musical training or sensitivity to a key change, would affect the stories people imagined. They analyzed people’s word choices using natural language processing, a branch of artificial intelligence where computers learn to understand text and speech.

The researchers determined the most commonly associated words and phrases for each listener, and mapped the stories onto a spatial representation of language, where words and sentences are placed closer together or farther apart based on how similar they are to one another.

Margulis and her team will use this data along with further studies to find out more about the factors that give rise to human creativity.

Music inspires memory

Understanding how music cues imagination and memory could open paths to treatment for Alzheimer’s disease and other forms of dementia, as well as inform other currently used applications of music therapy.

“There is a serious clinical potential to this kind of work,” Margulis said. “There are existing, well-supported clinical practices that ask people to recall memories while listening to music, but there hasn’t been a lot of research about the science around how these memories come up.”

Gabrielle Hooper, a graduate student in the Music Cognition Lab who studies the role of music in medicine, is conducting an analysis of the existing research on these topics. “By understanding how music impacts people with neurological diseases, we have the opportunity to introduce them to a new way to remediate their symptoms that may not have otherwise been available,” Hooper said. “For many dementia patients who experience a state of confusion at specific times of day, for instance, clinicians implement a personalized playlist to help ease their negative emotions.”

Margulis and a group of students in her lab are conducting studies that aim to link people’s autobiographical memories with the fictional stories that they imagine while listening to music. “We know from previous research that remembering something that happened to you relies on very similar mechanisms as imagining something totally fictional,” Margulis said. “We’re taking advantage of that relationship to bring together two lines of work about music that previously had not spoken to each other: narrative imagination and evoked memories.”

The team has conducted several studies aiming to understand which kinds of music are more likely to evoke memories, and which are more likely to cause someone to imagine a fictional story.

So far, they’ve found that in the same way people raised in similar cultures imagine similar stories, they also tend to recall similar memories when listening to music. “There are certain songs where people often tell us that they remember getting ready for prom with their friends,” Margulis said. “If you’re in a similar age demographic, there are specific ways that you were likely to hear music, which you’ll remember when it’s replayed.” Understanding these neural connections may lead to better understanding of how to implement music-based interventions in medicine.

Via her exploration of music through the lenses of music theory, psychology and cognition, Margulis has constructed new pathways for understanding music’s impact on our lives.

Her 2013 book, On Repeat, How Music Plays the Mind, digs into what features of music “hook” listeners, and how our craving for repetition is satisfied through music. In 2023, MIT Press published a book Margulis co-edited called The Science-Music Borderlands, which is a collection of interdisciplinary essays on topics in music psychology.

“Why is it that every known human culture has music?” Margulis said. “The answer that continually comes up is related to social bonding. If we’re listening to music, and we’re able to picture something similar, or remember something similar, then there is something there that connects us in an important way.”

Margulis is currently writing a new book that she hopes will inspire others to look at music through the intersection of science and art, and to harness the power of music to build bridges between disciplines and cultures.