Think of your favorite song. Maybe it brings you happiness or joy; maybe it makes you want to start dancing; or maybe it’s a sad, melancholic song that really speaks to you. Studies show that our brains release dopamine while listening to music, whether it’s Bach or Lady Gaga. This dopamine rush is the strongest when a song reaches its emotional climax and the listener feels the “chills” — the spine-tingling, hair-raising sensation of awe.

But what goes on inside the brain during those moments of musical delight?

Recent research by scientists at the University of California, San Francisco (UCSF) has revealed a small but profound secret of how something uniquely human like music moves us.

Zooming in on the auditory cortex, the researchers found three unique sets of neurons that light up while we listen to music. Two sets of neurons encode absolute pitch (individual musical notes) and pitch-change (the intervals between those notes). These circuits are shared when we also process speech.

The really interesting bit is the third set of neurons: these neurons only light up when we’re listening to music (and not speech) and try to predict what notes will come next, based on what notes you heard before.

“We found acoustic information, like pitch or the changes between pitches in a melody. But we also found information that was encoded that reflects the listener’s prior experience with music. This is learned information. It’s not bottom-up, it requires some internal model. Specifically, we found that listeners could predict the next note given the prior notes in the melody, and that information was encoded in a in a population of neurons in this auditory region,” Narayan Sankaran, a postdoc in the Chang Lab at UCSF and lead author of the study, told ZME Science.

“Interestingly, those neurons only responded to music. They didn’t respond to other sounds like speech or noise or the wind blowing or a cat. They only responded when a subject was listening to music,” he added.

Mapping Melody in the Brain

For Sankaran, the journey began with a question: What goes on inside the human brain when we listen to music to make us feel the way we do? Sankaran, once a musician himself, turned to neuroscience to explore how the auditory cortex transforms simple sound waves into rich emotional experiences.

“We studied epilepsy patients undergoing neurosurgery, who had electrode arrays implanted on the surface of their auditory cortex,” Sankaran explained. These rare circumstances provided an unprecedented view of the brain’s inner workings, while the patients listened to various melodic phrases from Western music, as well as sentences spoken in English. The 208 Western musical phrases varied systematically in pitch, pitch change, and expectation.

The team discovered three groups of neurons working in harmony. Two groups, already known to process pitch in speech, track the musical notes as they rise and fall. But the third group—a specialized set of neurons—predicts what note will come next. These neurons, Sankaran emphasized, “only responded when a subject was listening to music.”

These neurons explain what makes music appealing. If you were to isolate individual sounds from a song, there would be no emotional impact. But when we listen to a song from start to finish, we’re exposed to context. As we listen to a phrase, the brain encodes the statistical structure of music and tries to anticipate what notes will come up next. The best composers may not be neuroscientists, but they are masters at creating the right amount of tension and resolution to put these neurons to work.

“The more unpredictable or the more unexpected a note was, the larger the neural response in this music-specific population [of neurons]. If you encounter a highly unexpected note in a melody, these neurons really like to fire. If you’re just hearing a scale and everything’s very predictable, these neurons aren’t firing that much. So these music-specific pathways are really engaged when music is interesting and slightly unexpected,” Sankaran told me during an interview at the 2024 Falling Walls conference in Berlin.

Familiarity Matters

“The continuum between fulfilling and violating expectations is central to our aesthetic experience of music,” the authors wrote. This mechanism could also explain why certain songs evoke strong emotions or memories.

The findings suggest that while music and speech share some neural mechanisms for pitch and sequence processing, each has its own dedicated pathways. Speech-specific neurons, for example, encode the statistical structure of phonemes (the smallest unit of speech distinguishing one word), while music-specific neurons handle melodies. This is perhaps a good hint that these pathways share evolutionary roots. Why humans have the seemingly unique ability to process music, from an evolutionary standpoint, has been a hotly debated topic.

“Some people think that music and language started as the same thing as a proto-language. That’s a prominent hypothesis. To be honest, I don’t think we’ve tackled it empirically yet. We don’t have good data. It’s a hard, hard problem to tackle empirically. I think that music has conferred a lot of evolutionary advantages over time, so the ability to extract fine-grain patterns in our environment was something that we needed for our survival. Music just attached an element of pleasure to that process. So it was basically a way of doing an activity that was rewarding, but that also paid dividends for conferring some survival advantage. That’s personally what I believe the function of music has been, and in some ways continues to be,” Sankaran said.

The implications of this research extend far beyond understanding the basics of music perception. By decoding how melodies interact with the brain’s predictive and reward systems, scientists hope to harness music’s therapeutic potential. From treating PTSD and anxiety to enhancing cognitive function, music may prove to be an invaluable tool for healing.

Sankaran noted that people unfamiliar with Western tonal music might not engage in the same music-specific pathways as listeners steeped in its traditions. That’s an important limitation, which warrants more research to completely explore these music-encoding neurons.

This research focused on the melodic aspects of music. Looking ahead, Sankaran would like to focus on another vital dimension of music: timbre.

“We have two notes. One’s played by a trumpet, one’s played by a violin. They’re the same loudness, they’re the same pitch, they’re the same duration, but we can easily tell them the part. Why? Timbre. So I want to characterize how timbre is represented in this region [the superior temporal gyrus],” he said.

The findings were reported in the journal Science Advances.