Music and Emotions: How Sound Becomes a Shiver

Key takeaways

One well-studied route from sound to chills runs like this: fast brainstem reactions to the sound, expectations about what comes next, and dopamine release in the brain's reward system as the moment builds and peaks.

The minor third may sound sad because sad speech uses it too. In one study, actresses' sad speech tended to move by about a minor third. That link is suggestive, not proven.

Knowing a piece lets your brain anticipate its big moment, and reward regions ramp up in the seconds before it. Familiarity feeds the shiver.

Jazz improvisation may hit differently because you hear the expectations being set up and broken live. That's a reading of the theory, not a tested result.

The reward-system findings are solid and replicated. Prediction-based accounts like David Huron's ITPRA theory are strong frameworks still waiting on direct tests.

Why does a minor-key melody give you chills? Music and emotions research points to a chain of events: your brainstem reacts to the raw sound, your brain bets on what comes next, and your reward system releases dopamine as the moment builds and lands. This article walks that chain one link at a time, marks which links are measured and which are theory, then tests it on a jazz solo.

How are music and emotions connected, mechanically?

There isn't one route from sound to feeling. In an influential 2008 review, psychologists Patrik Juslin and Daniel Västfjäll listed six: brainstem reflexes, evaluative conditioning, emotional contagion, visual imagery, episodic memory and musical expectancy. For chills, two of those do most of the work. Brainstem reflexes respond to the raw sound, like a sudden loud entry or a harsh texture. Expectancy tracks where the music is heading and how much you want it to get there.

The dopamine part has been measured directly. In a 2011 Nature Neuroscience study, Valorie Salimpoor and colleagues used PET scans on eight listeners hearing music they had chosen themselves, and found dopamine release in the striatum, part of the brain's reward system, during peak pleasure. Brain scans of the same people then split the timing: the caudate was most active in the 15 seconds or so before a chill, and the nucleus accumbens at the chill itself.

So the felt emotion sits near the end of the chain, not the start. For the shorter version of the frisson story, start with The Shiver You Didn't Ask For.

What does a minor-key melody do to your brainstem before you 'feel' anything?

Your auditory system sorts sound before you can name what you're feeling. Recordings from the scalp show the brainstem already responds more strongly to consonant intervals than dissonant ones, in an order that mirrors how pleasant people rate them (Bidelman and Krishnan, Journal of Neuroscience, 2009).

Juslin and Västfjäll describe brainstem reflexes as fast and automatic: the system flags a sound as urgent before you think about it. That's one reason a sudden swell or a harsh chord can tighten your chest before you know why.

Roughness and the fear circuit

One acoustic feature gets special treatment. When a sound's loudness flickers very fast, roughly 30 to 150 times a second, the ear hears it as rough or harsh. In a 2015 Current Biology study, Luc Arnal and colleagues found that human screams sit in that range, along with alarms and dissonant intervals, while normal speech does not. The rougher a scream or alarm, the more alarming people rated it, and in brain scans both drove activity in the amygdala, a region central to processing fear.

That study tested screams and alarms, not music, and it says nothing about minor keys. A minor chord isn't especially rough. So the fair reading is narrower: harsh, dissonant textures in music borrow an acoustic signature your brain treats as a warning. Whether that warning helps a passage give you chills is a reasonable idea, not a measured result.

Why the minor third reads as sad

The sad sound of the minor key may come from speech. In a 2010 study in the journal Emotion, Meagan Curtis and Jamshed Bharucha recorded nine actresses speaking short two-syllable phrases in different emotions. In the sad versions, the two main pitches tended to sit about a minor third apart, and that interval was the most reliable cue listeners used to hear sadness.

That's one small study with trained actors, and some linguists doubt that everyday speech uses clean musical intervals at all. Broader work by Daniel Bowling and colleagues found that sad or subdued speech uses smaller pitch steps than excited speech in both English and Tamil, and that sad music in Western and South Indian traditions uses smaller steps too (PLoS ONE, 2012). That supports a link between how voices and melodies carry sadness. The authors are clear that it doesn't prove one causes the other.

Mapping which of these cues moves you is what Frisson is built for.

What does music and emotions research say about prediction and the reward loop?

Your brain is constantly predicting, and a melody is a series of bets. Juslin and Västfjäll list musical expectancy as one of the core routes from sound to feeling: when a phrase holds back the resolution you were counting on, tension builds, and the arrival registers as release.

Huron's ITPRA model in plain terms

Music psychologist David Huron's ITPRA theory, from his 2006 book Sweet Anticipation, splits your response to any expected event into five parts. Two come before it: Imagination (picturing what might happen) and Tension (getting ready for it). Three come after: Prediction (a quick reward or penalty for guessing right or wrong), Reaction (a fast, reflex-like response) and Appraisal (a slower, considered judgment).

Huron's account of chills follows from that. A surprise in the music sets off a fast, fear-like reaction, the kind that raises the hair on your arms. A moment later, the slower appraisal registers that you're safe and the surprise was good, and the contrast between the two feels like pleasure. That's why a deceptive cadence, a voice entering late, or a chord borrowed from another key can give you chills. A 2021 review of chills research by Rémi de Fleurian and Marcus Pearce found the evidence consistent with expectation playing a role, but said studies don't yet have the timing precision to confirm the exact sequence Huron proposes.

The chills circuit: anticipation to peak

The best timing data comes from Salimpoor's team. The caudate ramped up in the seconds before a chill. At the chill itself, activity peaked in the nucleus accumbens. Dopamine also looks causal, not just along for the ride. In a 2019 PNAS study, Laura Ferreri and colleagues gave 27 people a drug that boosts dopamine, a drug that blocks it, and a placebo on separate days. Pleasure ratings went up on the booster and down on the blocker, and the blocker made chills harder to get.

The amygdala's role is less clear. The first brain imaging study of music chills, by Anne Blood and Robert Zatorre in 2001, reported amygdala activity going down as chills grew stronger. So "tension in the amygdala, then reward in the accumbens" is a tidy story the data doesn't yet support.

If you play an instrument, you've felt this from the inside. We cover that side in What Playing an Instrument Does to Your Brain.

Why does jazz improvisation produce a different kind of emotional hit?

Here's how the expectation theory would read it. A composed melody plays with expectations you barely knew you had. A jazz solo plays with expectations you can hear being built, because you're listening to the musician make choices in real time. If expectation drives the chill, a setup you heard being built should make the payoff land harder. No study has tested that directly.

Case study: tracing one Coltrane chorus cue by cue

Take a Coltrane chorus over fast-moving changes. The chords shift every few beats, so your sense of where the harmony is going keeps resetting. Coltrane runs a phrase up the changes, and you sense where it should land. He overshoots it, or lands a half-step late, or answers his own phrase with a shifted echo instead of a repeat. In ITPRA terms, each of those is a broken prediction, followed by an appraisal that the note was right after all.

Syncopation as a prediction violation

Syncopation is the cleanest example. Your body locks onto the pulse quickly. The soloist then places notes off that pulse, and your expectation of when the next note should land gets broken. If the chill comes from the broken expectation, it should arrive at the moment of surprise, not on a predictable note. It's the kind of thing listeners struggle to put into words, the gap explored in The Song That Knows You.

What are the established facts about music and emotions studies?

Some findings hold up well.

Chills light up the reward system. Blood and Zatorre (PNAS, 2001) scanned 10 music students listening to music they chose because it reliably gave them chills. As chills grew stronger, activity changed in regions tied to reward, emotion and arousal, including the ventral striatum, midbrain and amygdala. The authors noted these same regions respond to food, sex and drugs of abuse. Salimpoor's 2011 study added direct evidence of dopamine release.

Most people get chills, but not everyone. In a week-long study that pinged people as they went about their day, 81% reported at least one music chill (Nusbaum et al., 2014). Earlier survey work they cite found about 8% of a student sample said they never get them. More on that in Why Do I Get Chills From Music?

Dopamine matters causally. Boosting or blocking it shifts how much people enjoy music (Ferreri et al., PNAS, 2019).

Other ideas are less settled. Huron's ITPRA theory is a framework, not a measurement, and the exact timing it proposes hasn't been confirmed.

Reward-region activity during chills: Solid, replicated imaging

Most people get music chills, some never do: Solid, surveys and diary studies

Dopamine involvement: Supported, including drug studies; timing still debated

Prediction as the main driver: Theory (ITPRA), consistent with the evidence but not directly tested

Minor third as a speech cue for sadness: One small study, suggestive

Where does your own answer sit? The songs people share on Frisson's leaderboard are personal stories, not data, and that's the honest gap between lived frisson and lab findings.

Why does the same piece give you chills and leave someone else cold?

Prediction runs on memory. Chill moments are usually tied to a passage you know well: the entry after a silence, the key change, the voice cracking. Salimpoor's data fit this, with the caudate ramping up in the seconds before a peak you know is coming. The authors note that anticipation could come from knowing the general rules of music, knowing the specific piece, or both. A first-time listener can still be caught by a surprise, but they have far less to anticipate.

Personal history stacks on top. The minor-key melody that gives you chills may be tangled up with a winter, a person, a first listen in a particular room. Juslin and Västfjäll list episodic memory as its own route to emotion for exactly this reason. That part lives in you, not in the sound.

This is why the shiver resists a single explanation. It's your fingerprint: the same chord progression, filtered through your training, your memories and your expectations. The piece is the trigger. The response is yours.

What do quotes about music and emotions get right and wrong?

"Music is the language of emotion" holds up better than its critics admit. Sad speech and sad melodies share features, like smaller pitch steps, in at least English and Tamil, and one study found sad speech leaning on the minor third. So a quote about a sad melody may be describing a real acoustic signal, not just projection.

Where quotes fail is causation. "Music speaks what words cannot" implies emotion flows straight from sound. The research points to several routes running in parallel: fast reflexes, expectation, memory and reward. Quotes also flatten individual differences. A line about chills can't explain why one piece moves you and leaves a friend cold. Treat quotes as accurate descriptions of the destination and unreliable maps of the route.

How can you map your own emotional response to music?

Start with the piece that reliably gives you the shiver. Listen twice: once to catch the moment the chill arrives, once to isolate what happens just before. Chills often follow something unexpected, so the trigger tends to sit a beat or two before the feeling. Write down what you hear: a swell into a new key, a voice re-entering, a solo climbing past where the harmony said it would stop. Do that across three or four trigger pieces and you have the start of your own fingerprint.

What Frisson asks, and what it cannot tell you

Frisson (tryfrisson.com, free, no password needed) asks one question about the song that moves you and reads your answer for the underlying pattern. It can name the shape of your response. It can't scan your brain or run the chain for you; that stays in the lab. What you get is vocabulary precise enough that the next time someone asks why that one passage tightens your chest, you can point to the bar instead of shrugging.

FAQ

Why do I get goosebumps from one specific song but not others?

Chills ride on expectation, and expectation rides on memory. A trigger usually sits at a moment you know well: the entry after silence, the key change, the crack in a voice. Brain imaging shows reward regions ramping up in the seconds before a peak you know is coming. Personal associations stack on top of that.

Is frisson the same as ASMR?

No. Frisson is a quick wave of chills tied to an emotional moment in music, and it tends to spread fast across the body. ASMR is a tingling that usually starts at the head and neck, is set off by sounds like whispering or tapping, comes with relaxation, and can last for minutes. Both are pleasant bodily sensations, but their triggers and time courses differ.

Does sad music actually make you sad, or something else?

Partly. Sad melodies share acoustic features with sad speech, so your brain can read them a bit like a sad voice. But what people feel is often more mixed: being moved, tension and release, even pleasure. An analysis of about 1,000 pieces that give people chills found they lean sadder in tone than average, which suggests sadness and pleasure often travel together.

Can you train yourself to feel music more intensely?

Possibly, though no study has shown a training method that works. Since chills ride on expectation, knowing a piece well gives your brain more to anticipate. You can also train attention: listen once for the moment the chill arrives, and once for what happens a beat or two before. Tracked across several trigger pieces, that pattern becomes your fingerprint.