In 1993, neuroscientist Jaak Panksepp published an unusual survey. He asked respondents whether music had ever made them feel chills — a shiver or tingling sensation running up the spine or across the scalp. Almost all said yes. He then asked which types of music produced this response most reliably.

The answers were strikingly consistent: slow tempos, unexpected harmonic shifts, swelling dynamics, a soprano voice sustaining a high note, the moment a new instrument enters a passage already dense with tension. These were not arbitrary preferences. They pointed toward something structured — a set of musical features that reliably triggered a specific physiological response across diverse listeners.

Panksepp was fascinated by this because it was so biologically anomalous. The chill response — piloerection, the same mechanism as goosebumps in cold — is an ancient mammalian reflex. In other mammals, it serves a purpose: fur bristling makes animals appear larger to predators, or responds to sudden cold. In humans responding to music, it appears to serve no purpose at all. It is the body reacting to sound with a response evolution built for physical survival.

What is happening? And why should vibrations in air produce tears, joy, chills, or the almost unbearable feeling of being understood?

The answers, emerging from neuroscience over the last thirty years, turn out to involve fundamental questions about prediction, reward, memory, and what it means to be a brain that makes sense of the world by anticipating it.

"Music is the shorthand of emotion." — Leo Tolstoy

Key Definitions

Frisson — The "aesthetic chill" or "musical shiver" response: piloerection, tingling, brief acceleration of heart rate, sometimes tears, in response to particularly moving musical moments. Reported by approximately 50-70% of people. Associated with dopamine release in the nucleus accumbens.

Dopamine prediction error — The neural signal generated when an outcome differs from what was predicted. Positive prediction error (better than expected) produces reward learning; negative prediction error (worse than expected) produces aversion. Huron's theory proposes musical emotion is largely driven by prediction errors in musical expectation.

Musical expectation — The brain's implicit predictions about where a piece of music is going — the next note, the harmonic resolution, the rhythmic continuation — based on learned musical patterns. Emotion arises from the fulfillment, violation, or delay of these expectations.

SEEKING system — Jaak Panksepp's term for the mammalian motivational system that produces anticipatory arousal, curiosity, and the drive toward rewarding experience. Mediated by dopamine; proposed as the neural substrate through which music activates intense emotional states.

Entrainment — The synchronization of neural oscillations (and body movement) with an external rhythmic stimulus. The brain literally oscillates in time with music, and bodies naturally move to synchronize with rhythmic beats.

Consonance and dissonance — Perceptual qualities of pitch combinations: consonance is the pleasing quality of simple frequency ratios (perfect fifth: 3:2; octave: 2:1); dissonance is the tense, rough quality of complex frequency ratios. Involves both acoustic physics and culturally learned expectations.

Musical absolute pitch (perfect pitch) — The ability to identify or produce a musical note without a reference pitch. Present in approximately 1 in 10,000 of the general population; much higher among those with early musical training; associated with specific structural differences in auditory cortex.

The Prediction Machine Listens to Music

The modern neuroscience of music is inseparable from the modern theory of the predictive brain.

Karl Friston, Andy Clark, and others have developed the view that the brain is fundamentally a prediction machine — continuously generating models of what will happen next, comparing those predictions to actual sensory input, and updating the model based on prediction errors. Perception is not passive reception of the world; it is active hypothesis-testing.

Music activates this prediction machinery intensively. After even brief exposure to a musical tradition, the brain has internalized its patterns: the scales, the chord progressions, the rhythmic conventions, the melodic contours that characterize the style. When you listen to a melody, your brain is simultaneously perceiving the notes and predicting what will come next — generating anticipation.

David Huron's Sweet Anticipation (2006) is the most comprehensive theory of how this prediction system generates musical emotion. Huron identifies several distinct emotional systems activated by musical expectations:

Imagination: Simply imagining a musical continuation — before the music plays — activates reward circuits. The anticipation of a coming resolution creates pleasurable arousal.

Tension: Sustained uncertainty about resolution — the drawn-out dominant seventh chord that hasn't resolved yet, the silence before the final chord — creates physiological arousal and subjective tension.

Prediction: The moment-by-moment accuracy (or inaccuracy) of musical prediction generates continuous reward-and-surprise signals.

Reaction: The immediate physiological and emotional response when an expected event occurs (or fails to occur). Frisson occurs here — at the moment of surprise, the unexpected beauty, the particularly perfect resolution.

Appraisal: The post-hoc cognitive evaluation of what just happened — "that was beautiful," "that was unexpected," "that moved me."

The Neuroscience of Frisson: Dopamine and Expectation

Anne Blood and Robert Zatorre at McGill University conducted the first neuroimaging study of musical frisson in 2001. Participants brought music that reliably gave them chills; the researchers scanned their brains while they listened.

The results were striking. Moments rated as producing chills activated the nucleus accumbens, ventral tegmental area (VTA), amygdala, insula, and orbitofrontal cortex — the core reward circuit that responds to food, sex, addictive drugs, and other biological rewards.

This meant music was triggering genuine dopamine release in the brain's reward system through purely abstract, auditory stimulation. Nothing was being ingested; no biological need was being met; no survival-relevant event had occurred. Sound patterns were activating one of evolution's most ancient and fundamental motivational systems.

Valorie Salimpoor and colleagues (2011) extended this work with a more direct measure. Using radiolabeled raclopride (which binds to dopamine receptors — less binding means more dopamine release, since endogenous dopamine and the tracer compete for the same receptor), they found that chills-producing music released dopamine not just at the peak moment but in the anticipation period leading up to it — consistent with the prediction error model (anticipatory dopamine for expected reward; confirmation dopamine at delivery).

The anatomy was telling: dopamine released in the caudate during anticipation; dopamine released in the nucleus accumbens at the chills moment itself. The caudate-nucleus accumbens distinction maps onto the wanting-vs-getting distinction in reward neuroscience — the brain responds differently to anticipating a reward and actually receiving it.

Why Not Everyone Gets Musical Chills

Roughly 30-50% of people report never experiencing frisson from music. This is not a failure of sensitivity or taste. Matthew Sachs and colleagues found a structural brain difference: people who experience frisson have a denser connection between the auditory cortex and the regions involved in emotional processing, including the anterior insula and the medial prefrontal cortex.

This suggests that frisson proneness partly reflects the richness of the structural pathway through which auditory information reaches emotional processing systems. More axons, faster communication, richer emotional resonance.

Individual differences in openness to experience — the Big Five personality trait — also predict frisson proneness. Open individuals seek novelty, aesthetic experience, and variety; they are more likely to be deeply engaged by musical surprises and to experience their emotional impact fully.

Music and Memory: Why Songs Take You Back

Music is the most powerful autobiographical memory cue known to psychology.

David Rubin's research on memory for music found that musical memories — compared to memories for other stimuli — show unusual vividness, emotional intensity, and involuntary recall. When a piece of music associated with a specific autobiographical period is heard, it can retrieve not just the memory but the full emotional context: the feeling of being 17, the specific quality of a relationship, the emotional tenor of an entire season of life.

The neural basis involves the hippocampus, which encodes autobiographical memories, and the limbic system, which stores the emotional context. The temporal lobe's role in music processing overlaps substantially with autobiographical memory networks; hearing a piece of music activates both the acoustic representation and the contextual memory network in which it was encoded.

This explains why music heard repeatedly during formative emotional experiences becomes permanently associated with those experiences — and why hearing it later retrieves the original emotional state with unusual fidelity.

The Reminiscence Bump

The well-documented reminiscence bump in autobiographical memory — the tendency for adults to recall disproportionately many memories from ages 15-25 — is particularly pronounced for music. Music that was popular during that period retains unusual emotional salience throughout life.

Stefan Janata at UC Davis has shown that the medial prefrontal cortex, active during self-referential processing and autobiographical memory retrieval, shows greater activation to personally familiar music than to unfamiliar music — and this personal familiarity effect is stronger the more emotionally significant the association.

This is clinically relevant: music therapy for Alzheimer's disease can reactivate memories and emotional states in patients with severe semantic memory loss, because the emotional memory system is partially preserved even when other memory systems are damaged. Patients who no longer recognize family members may still respond emotionally to music from their youth.

The Social Dimension: Why Music Is Always Cultural

Music is a universal human feature — no known culture lacks it. But music is also always cultural: every human musical tradition creates its own distinctive system of expectation, its own vocabulary of tension and release, its own grammar.

The emotional power of music is partly universal and partly constructed by cultural learning.

What Appears Universal

Patrick Savage's cross-cultural comparative musicology and Samuel Mehr's large-scale study across 60 societies find consistent cross-cultural features:

  • Rhythm and beat: Rhythmic entrainment — synchronizing body movement to a beat — appears universal; no culture has been found without rhythmic music
  • Arousal recognition: Fast tempo = high energy; slow tempo = low energy — recognized across cultures
  • Functional categories: Lullabies (soothing, simple, descending), dance music (rhythmic, energetic), and healing/ritual music (often involving repetition and drone) are distinguishable across cultures
  • Sad music universals: Slow tempo, minor mode (in traditions using it), low pitch, soft dynamics — these features cluster in "sad" music across independent traditions

What Varies

  • The specific scales, intervals, and harmonic progressions that feel "resolved" vs. "unresolved"
  • The emotional valence of specific timbres and modes
  • The boundary between music and speech/chant
  • Whether specific emotional categories (nostalgia, triumph, yearning) are musically encoded and recognized

The Moving Body: Why We Can't Stop

One of music's most universal effects is involuntary movement. The urge to tap a foot, nod a head, or dance is nearly impossible to fully suppress.

This reflects the deep integration of music processing with the motor system.

Neural entrainment: When the brain processes rhythmic music, neural oscillations in auditory cortex synchronize with the musical beat — the auditory system literally oscillates in time with the music. These oscillations propagate to motor regions, producing the motor preparation that becomes movement.

The basal ganglia and cerebellum: Both structures are involved in both rhythm processing and motor control. Basal ganglia lesions (as in Parkinson's disease) disrupt both voluntary movement initiation and rhythmic entrainment. Notably, rhythmic auditory stimulation (heard rhythmic music) can temporarily improve gait in Parkinson's patients — a documented clinical application of musical rhythm's grip on motor systems.

The universality of dance — moving the body in rhythmic synchrony with music or other people — appears to reflect this deep motor-auditory integration, and has been proposed as serving social cohesion functions: synchronized movement between individuals produces feelings of affiliation, cooperation, and trust.

Why Music Can Hurt As Much As It Helps

The same properties that make music emotionally powerful make it difficult to control.

Earworms (involuntary musical imagery, or INMI) — the experience of a fragment of music playing repetitively in one's mind — affects nearly all people, typically for minutes at a time, occasionally for hours or days. James Kellaris has studied INMI extensively; it is most common for music with simple, repetitive patterns, unexpected intervals, or strong associations with recent contexts. The brain's prediction system, having set up strong musical expectations, continues generating the anticipated continuation even in the absence of the music.

Emotional induction: Music can induce emotional states that persist after the music ends — including sad emotional states. Listening to genuinely sad music (slow, minor, descending) can produce sad affect that lingers. In people already in sad states, sad music may deepen rather than relieve the sadness — consistent with mood-congruent processing. Research on music in clinical depression is mixed: some studies find music therapy beneficial; others find that melancholic music preferred by depressed individuals may maintain rumination.

The autonomic signature: Music-induced emotion produces genuine autonomic arousal — changes in heart rate, skin conductance, respiration rate, and body temperature — that can be measured objectively. Stefan Koelsch's meta-analysis documented these autonomic signatures across studies, confirming that musical emotion is not simply a cognitive report but a full-body physiological response.

Music and the Developing Brain

Musical training beginning before age 7 produces the most extensive brain plasticity effects. The auditory system during this window is in a sensitive period — maximally responsive to acoustic experience.

Compared to non-musicians, musically trained individuals show:

  • Enlarged primary auditory cortex with more refined frequency mapping
  • Enhanced brainstem encoding of acoustic features (the auditory brainstem response is faster and more precise)
  • Enlarged corpus callosum (particularly the anterior portion, connecting frontal and temporal areas involved in music)
  • Enhanced verbal memory and phonological awareness
  • More efficient attentional control

Nina Kraus's lab at Northwestern has documented the brainstem effects extensively, showing that musicians' nervous systems encode acoustic features more faithfully even outside the cortex — suggesting music training's effects are not superficial but reach the deepest levels of auditory processing.

The reason music moves us turns out to be neither mysterious nor accidental. It engages the prediction systems that make the brain work; it activates reward circuits that evolution built for survival; it retrieves memories with unusual fidelity; it synchronizes bodies through motor-auditory integration that is millions of years old.

What is remarkable is not that music affects us so deeply. What is remarkable is that we ever thought it was merely entertainment.

For related articles, see why we cry, how learning happens in the brain, what causes anxiety, and why exercise feels good.

References

  • Blood, A. J., & Zatorre, R. J. (2001). Intensely Pleasurable Responses to Music Correlate with Activity in Brain Regions Implicated in Reward and Emotion. PNAS, 98(20), 11818–11823. https://doi.org/10.1073/pnas.191355898
  • Salimpoor, V. N., et al. (2011). Anatomically Distinct Dopamine Release During Anticipation and Experience of Peak Emotion to Music. Nature Neuroscience, 14(2), 257–262. https://doi.org/10.1038/nn.2726
  • Huron, D. (2006). Sweet Anticipation: Music and the Psychology of Expectation. MIT Press.
  • Panksepp, J. (1995). The Emotional Sources of 'Chills' Induced by Music. Music Perception, 13(2), 171–207. https://doi.org/10.2307/40285693
  • Sachs, M. E., et al. (2016). Brain Connectivity Reflects Human Aesthetic Responses to Music. Social Cognitive and Affective Neuroscience, 11(6), 884–891. https://doi.org/10.1093/scan/nsw009
  • Mehr, S. A., et al. (2019). Universality and Diversity in Human Song. Science, 366(6468). https://doi.org/10.1126/science.aax0868
  • Kraus, N., & Chandrasekaran, B. (2010). Music Training for the Development of Auditory Skills. Nature Reviews Neuroscience, 11(8), 599–605. https://doi.org/10.1038/nrn2882
  • Janata, P. (2009). The Neural Architecture of Music-Evoked Autobiographical Memories. Cerebral Cortex, 19(11), 2579–2594. https://doi.org/10.1093/cercor/bhp008

Tags

music neuroscience emotion dopamine psychology brain creativity

Frequently Asked Questions

Why does music give you chills or goosebumps?

The chills or goosebumps response to music is called 'frisson' (from the French for 'shiver'), or sometimes 'musical chills' or 'aesthetic chills.' It is a genuine physiological response — piloerection (the same mechanism as goosebumps in response to cold), often accompanied by a tingling sensation on the scalp, spine, or extremities, brief elevation in heart rate, and in some cases tears. Approximately 50-70% of people report experiencing frisson in response to music at some point, though frequency and intensity vary. Anne Blood and Robert Zatorre's 2001 PET study was the first to demonstrate the neural basis: music that participants rated as producing 'chills' activated the nucleus accumbens (the brain's primary reward hub) and the ventral tegmental area (VTA) — the same dopaminergic reward pathway activated by food, sex, and addictive drugs. This suggested that musical frisson involves genuine dopamine release in response to purely abstract, auditory stimulation — not to any biological reward. David Huron's theory of musical expectation provides the mechanism: music creates expectation of where it will go next (through established patterns, harmonic rules, rhythmic momentum), and frisson occurs at moments of surprise, unexpected resolution, or particularly beautiful fulfillment of expectation. The listener's brain has predicted a musical outcome; the actual outcome is better than expected or different in a moving way; the dopamine prediction error system fires. Neuroscientist Jaak Panksepp proposed that musical chills engage the brain's 'SEEKING' system — the motivational system that produces anticipatory arousal — and that the emotional power of music is linked to the arousal of this ancient mammalian motivational system.

Why does music make people cry?

Emotional responses to music — including tears — are nearly universal across cultures and appear from early infancy, suggesting they are not purely learned responses but engage deep emotional systems. The specific triggers and mechanisms of musical weeping have been studied by Tuomas Eerola, Suvi Saarikallio, and others. Several distinct pathways from music to tears have been identified. Expectation violation and resolution: as with frisson, moments where musical expectation is subverted or beautifully resolved can produce profound emotional responses. Sudden major-to-minor shifts, unexpected harmonic resolutions, and melodic rises followed by falls are reliably moving. Association and memory: music is a powerful autobiographical memory cue (research by David Rubin and others shows music-evoked memories are particularly vivid and emotionally intense). A piece of music associated with a lost relationship, a deceased person, or a formative life moment can trigger the full emotional context of those memories, including grief. This is the mechanism behind the common experience of a song bringing tears unexpectedly — the music has retrieved a memory with its emotional freight intact. Prosodic mirroring: music shares structural features with emotional vocalizations — pitch contour, tempo, timbre. Slow, descending pitch contours and falling dynamics appear in both sad music and sad speech across cultures. The auditory system may process these as genuine emotional communications, activating emotional responses directly. Perceived social emotion: people cry at music partly in response to perceiving the emotion expressed by the music or the performer — an empathic response to the perceived emotional state encoded in the sounds. Context matters: the same musical passage is more likely to produce tears in a concert hall with shared social context than heard through headphones alone.

Does every culture respond to music the same way emotionally?

This is one of the most important and contested questions in the psychology of music. The dominant assumption in Western research has been that certain musical features have universal emotional meanings — minor keys are sad, major keys are happy, fast tempos are energetic and joyful. The first major challenge came from Samuel Mehr and colleagues' (2019) study involving 60 societies with recordings of songs from each. Participants across cultures could identify whether songs were for dancing, soothing infants, healing/ritual, or expressing love — suggesting some universal recognition of musical function. But the picture is more nuanced for specific emotions. Patrick Savage and colleagues' cross-cultural work found some universal features — tempo, rhythm, roughness/smoothness of timbre — but substantial cultural variation in how specific musical features are mapped to emotions. A 2020 study by Samuel Mehr's lab (published in Current Biology) playing Western and tribal music to isolated indigenous Mbenzele Pygmies in the Central African Republic found they could identify relaxing vs. energizing songs from Western music above chance, suggesting some cross-cultural recognition of basic arousal dimensions. However, the specific emotion labeling of music — distinguishing sadness from fear from anger, for instance — appears significantly culturally mediated. The Balinese response to music Western listeners find 'threatening' or 'sad' is not necessarily congruent. The broadly universal elements appear to be: recognition of energy level (arousal), some recognition of valence extremes (very positive vs. very negative), and recognition of music intended for specific social functions (lullabies, dance music). The specific, finely differentiated emotion labels (nostalgic, triumphant, yearning) appear to require cultural learning.

What happens in the brain when you listen to music?

Music activates an unusually extensive network of brain regions — more so than almost any other natural stimulus. This reflects the multidimensional nature of music: it involves auditory processing, rhythm synchronization, emotional processing, memory retrieval, prediction and expectation, motor preparation (even without movement), and in many contexts, social communication. Key regions and their roles: The auditory cortex (primary and secondary) processes the basic acoustic features — pitch, timbre, temporal patterns, harmonic relationships. The cerebellum and basal ganglia process rhythm and synchronize neural oscillations with the musical beat — these regions activate even in completely deaf people exposed to beat via vibration, suggesting rhythm processing has ancient, pre-auditory roots. The motor system — including pre-supplementary motor area and M1 — activates during music listening even without movement, explaining the irresistible urge to move to rhythm and the phenomenon of rhythmic entrainment (synchronizing body movement to musical beat). The limbic and paralimbic regions — amygdala, anterior insula, anterior cingulate cortex, and ventral striatum — process the emotional dimensions. The hippocampus activates in response to familiar music and music associated with autobiographical memories. The default mode network activates during deeply engaging music listening, suggesting a self-referential, inward, imaginative dimension to musical experience. fMRI studies of professional musicians show expanded and more efficient auditory and motor cortex representations, supporting the plasticity claim that musical training literally reshapes the brain.

Why do we find some music beautiful and other music grating — is taste subjective or objective?

Musical aesthetics — why certain sounds are experienced as beautiful and others as unpleasant — involves both universal features and culturally learned preferences. On the universal side: consonance (the pleasing quality of simple frequency ratios — perfect fifths, octaves, major thirds) vs. dissonance (complex, 'rough' frequency ratios) appears to be partly determined by the mathematics of the harmonic series and the physiology of the auditory system. Josh McDermott's cross-cultural research in the Bolivian Amazon with the Tsimane people — who had minimal exposure to Western music — found that the Tsimane showed reduced preference for consonance vs. dissonance compared to Westerners, suggesting that Western consonance preferences are partly learned and partly universal. Cultural learning is clearly powerful: the scale systems, rhythmic conventions, and harmonic languages of different musical traditions create expectations against which music is evaluated. Indian classical music, Arabic maqam, and Western tonal harmony all create beauty through the manipulation of expectations within their respective systems. Listeners unfamiliar with a tradition may find its departures from expected norms grating (hearing dissonance where a tradition hears beauty) because their auditory system has not internalized the tradition's expectations. Familiarity itself increases liking (the mere exposure effect) — music that initially sounds strange often becomes appealing with repeated listening as the brain extracts its patterns. The general principle: beauty in music appears to arise from the right balance of predictability and surprise — enough familiarity to generate expectation, enough novelty to fulfill and subvert those expectations in stimulating ways. This is consistent across cultures, though what counts as familiar and surprising depends heavily on cultural musical experience.

Why does listening to music while working help some people and hurt others?

The effect of background music on cognitive performance depends on the complexity of the task, the nature of the music, and individual differences in personality and cognitive style. The Arousal and Mood Hypothesis, proposed by Sarnoff Mednick and developed by Nick Perham and others, suggests that background music affects performance primarily through its effects on arousal level and mood, which in turn affect cognitive processing. For simple, repetitive tasks (data entry, assembly work, routine physical tasks), moderately arousing background music improves performance and mood, consistent with Yerkes-Dodson: these tasks require some arousal but not high cognitive load, and music can supply arousal without competition. For complex cognitive tasks — particularly those requiring language processing (reading with comprehension, writing, learning new verbal material) — music with lyrics is reliably distracting, because the language systems that process lyrics compete with the language systems engaged in the task. Music without lyrics is less disruptive to verbal tasks. For tasks requiring deep concentration and working memory — complex problem-solving, learning new material, creative synthesis — most background music is distracting regardless of lyrics, competing for the attentional and working memory resources the task requires. Individual differences: extroverts generally benefit more from background music stimulation (being below their optimal arousal baseline in quiet environments) while introverts, already at or above optimal arousal, tend to show impaired performance with background music. People high in neuroticism are more vulnerable to distraction by background music. The practical recommendation: use music strategically — for mood and arousal during routine tasks, but remove it during tasks requiring sustained focused cognition.

Can music actually change the brain — are there lasting effects from musical training?

Musical training produces some of the most dramatic and well-documented examples of experience-dependent brain plasticity. The evidence is extensive and spans structural, functional, and cognitive changes. Structural changes: Nina Kraus, Stefan Koelsch, and others have documented that musically trained individuals show enlarged primary auditory cortex, larger corpus callosum (facilitating interhemispheric communication, particularly important for the bilateral demands of music), more extensive motor cortex representation of the hands (particularly in string players), and in pianists, specific enlargement of the cortical representation of the fingers. The amount of change correlates with years of training and is larger when training began before age 7 (consistent with early sensitive period effects). Auditory processing advantages: Musicians show superior pitch discrimination, enhanced ability to hear speech in noise (the 'cocktail party effect'), better encoding of rhythmic patterns, and more precise neural tracking of acoustic features. Kraus's work has shown that these advantages extend to speech processing and reading, suggesting music training may benefit literacy. The 'musician advantage' in noise: musicians' brainstems show more robust auditory encoding, detectable in the auditory brainstem response (ABR) — a millisecond-scale neural response to sound. Cognitive transfer: decades of research have found modest but consistent transfer effects from musical training to non-musical domains — verbal memory, executive function, phonological awareness. Whether musical training causes these cognitive advantages or whether children with these capacities are more likely to continue music training remains partially debated. Experimental studies assigning children to music training (rather than selecting already-musical children) find genuine transfer effects to reading and verbal memory, supporting causal claims.

Share this article