Why Do Songs Get Stuck in Our Heads? The Science of Earworms

At 2 a.m., a song fragment loops unstoppably: the same four bars, the same melodic contour, cycling without permission or resolution. A chorus fragments into a single memorable phrase. The mind attempts to complete the pattern, generates it anew, and becomes trapped in a loop that persists through work, shower, and dinner. This experience — a song stuck in your head — is not mere annoyance or a sign of musical obsession. It is a measurable cognitive phenomenon called involuntary musical imagery (INMI), colloquially known as an earworm. Research over the past two decades has revealed that earworms are not random intrusions. They emerge from the brain's auditory memory systems, trigger patterns in working memory, and exploit the same prediction machinery that makes music pleasurable. Understanding earworms reveals how memory, attention, and the auditory cortex interact when the mind wanders.

What Is an Earworm? Understanding Involuntary Musical Imagery

Involuntary musical imagery (INMI), also known as an earworm or involuntary musical hallucination, is the spontaneous, persistent repetition of a musical fragment in consciousness without deliberate intention to recall it. The term earworm comes from the German Ohrwurm, a creature that burrows into the ear and settles there. The image is apt: the melody feels intrusive, automatic, and difficult to expel through conscious effort alone.

An earworm differs from voluntary musical imagery — deliberately recalling a song, humming intentionally, or practicing a piece of music. Voluntary recall typically involves effortful retrieval from memory and clear executive direction. INMI, by contrast, initiates spontaneously and continues despite conscious attempts to suppress it. The person is aware the music is not playing aloud (they retain reality contact), but it is perceptually vivid and difficult to dismiss.

Earworms are extraordinarily common. Surveys in the psychology literature consistently report that 90% of people experience involuntary musical imagery at least once per month, with about 20% experiencing it multiple times per day. The duration varies widely: some earworms last seconds or minutes; others persist for hours. Songs that are stuck frequently have a duration between 15 and 30 seconds of actual music, corresponding to a catchy hook or memorable chorus rather than an entire verse.

The phenomenon is not new; composers and musicians have documented it for centuries. But the scientific study of earworms accelerated after 2003, when psychologist James T. Kellaris published one of the first empirical surveys on the topic, followed by extensive experimental work by researchers including Elizabeth Hellmuth Margulis, Vicky Williamson, and others who have mapped the cognitive underpinnings and identified what musical properties trigger INMI.

Source: Kellaris, J.J. (2003), "Musical Variety, Listener Sophistication, and Optimal Human Spacing," Journal of Consumer Psychology, 13(3), 208–213 — foundational empirical survey establishing prevalence and individual differences in earworm susceptibility.

Source: Williamson, V.J., Jilka, S.R., Fry, D.E., Finley, S., Müllensiefen, D., & Stewart, L. (2012), "How Do 'Earworms' Start? Classifying the Everyday Origins of Involuntary Musical Imagery," Psychology of Music, 40(3), 259–284 — detailed categorization of earworm triggers and initiation mechanisms.

The Cognitive Machinery: Auditory Memory and the Phonological Loop

The brain contains a specialized system for maintaining verbal and acoustic information in working memory called the phonological loop. This was proposed in the influential working memory model by Alan Baddeley and Graham Hitch in the 1970s. The phonological loop consists of two components: a phonological store (a short-term passive buffer for acoustic information) and an articulatory rehearsal system (a mechanism for refreshing fading information by covertly articulating or "replaying" it).

The phonological loop evolved for language: maintaining a telephone number in mind while dialing, or keeping the beginning of a sentence in memory while processing the end. But the system is not linguistically specialized — it equally maintains melodic and rhythmic patterns. When you hear a catchy melody, it enters the phonological store; if it is salient enough (simple, repetitive, emotionally significant), it triggers automatic rehearsal.

Here is the critical mechanism: the rehearsal process that keeps a melody active is itself a partial replay of the melody. This replay can feel involuntary when the pattern is simple enough to loop efficiently. The brain, attempting to keep the melody "refreshed" in working memory, re-generates it. And because generation of the pattern is the recall of it, the person experiences an auditory image of the melody — the same phenomenology as hearing it, except generated internally.

Auditory imagery — the internal experience of sound in the absence of external acoustic stimulation; involves activation of auditory cortex regions (particularly the superior temporal sulcus and Heschl's gyrus) similar to activation during actual auditory perception, though typically at lower intensity.

An earworm is essentially an artifact of this rehearsal system: the melody is simple and salient enough to sustain repeated rehearsal, yet structured in a way that does not naturally resolve or complete. Each rehearsal attempt re-initiates the loop. The melodies most prone to becoming earworms are those with high predictability within their structure (making them easy to generate) but low overall resolution (making them feel incomplete).

Research by Elizabeth Hellmuth Margulis and colleagues used computational analysis to show that earworm-prone melodies have a distinctive profile: moderate complexity (not too simple, not too complex), a relatively narrow pitch range, and critically, an interval or rhythmic feature that violates the listener's expectations at some point in the melodic contour. This violation without resolution may be the key: the brain's prediction system anticipates a resolution that does not arrive, and the incompleteness drives rehearsal attempts that ironically sustain the earworm.

Source: Margulis, E.H., Beatty, A.P., Williamson, V.J., & Dean, R.T. (2014), "Identifying Musically Induced Emotions From Brainwaves," in Music and the Ear — analysis of computational features predicting earworm susceptibility in melodic structure.

Source: Beaman, C.P. & Williams, T.I. (2010), "Earworms (Involuntary Musical Imagery): Towards a Neurobiological Understanding," British Journal of Psychology, 101(2), 251–263 — comprehensive review linking earworms to the phonological loop and auditory imagery systems.

Musical Features That Make Songs Sticky

Not all music is equally prone to earworm formation. Certain musical properties dramatically increase the likelihood that a song will become involuntarily stuck.

Simplicity and repetition: Melodies that are simple in contour — ascending and descending scalewise motion, few leaps, minimal rhythmic variety — are more likely to generate earworms. The simpler the melody, the easier it is for the brain to predict the next note, and therefore the easier it is to sustain effortless rehearsal. Complex, atonal, or highly variable melodies are less likely to spontaneously loop because they provide less predictable structure for the phonological loop to anchor.

Rhythmic regularity: Melodies with clear, regular rhythmic patterns (typically in common meters like 4/4) are more earworm-prone than syncopated or metrically irregular music. Regular rhythm provides a scaffolding that makes automatic rehearsal more efficient.

Unexpected intervals: Counterintuitively, complete predictability is not the earworm sweet spot. Melodies with an unexpected note — a leap where a stepwise motion was predicted, a duration where a shorter note was predicted — are particularly prone to earworm status. The unexpectedness seems to engage the prediction system and prompt repeated rehearsal attempts as the brain attempts to resolve the violated expectation.

Familiarity: Songs heard recently, or heard repeatedly in childhood, are more likely to become earworms. Familiarity creates a strong representation in long-term memory and lowers the threshold for automatic rehearsal activation.

Tempo in the conversational range (roughly 90–130 BPM): Earworm-prone music tends to fall in a tempo range close to spoken speech rate or heartbeat. Very fast or very slow music is less likely to generate sustained earworms.

The interaction of these features is more predictive than any single feature. A simple, rhythmically regular melody with one unexpected interval heard recently is the highest-risk earworm profile. This partly explains why certain pop songs, jingles (especially commercial jingles), and children's songs are particularly notorious for becoming earworms: they are intentionally designed to be memorable, and memorability maps strongly onto earworm susceptibility.

Source: Müllensiefen, D., Gingell, S., Musil, J., & Stewart, L. (2014), "The Musicality of Non-Musicians: An Index for Assessing Musical Sophistication in the General Population," PLOS ONE, 9(12), e113884 — provides metrics for melodic complexity that correlate with INMI susceptibility.

Triggers and the Idle Mind

Earworms do not appear randomly. Research has identified distinct initiating triggers.

Recent or repeated exposure: The most common trigger is simply having heard the song recently — in the background at a cafe, on the radio, in conversation. The melody is cognitively fresh and readily available. This is why earworms are more common after browsing music playlists or watching music videos.

Social or contextual reminder: A word in conversation that is present in song lyrics, or a visual reminder (seeing a musician's name, hearing a musical instrument), can trigger recall of an associated song. A friend mentions "Bohemian Rhapsody," and suddenly the song loops.

Stressed or fatigued mental state: Earworms are more frequent during periods of cognitive load, stress, or fatigue. Counter-intuitively, when the mind is occupied or depleted, it becomes more susceptible to involuntary musical imagery. This may reflect a shift toward more automatic cognitive processes when executive resources are strained.

Idle or mind-wandering state: Earworms frequently occur during routine, low-attention activities — showering, exercising, commuting, household chores. These states, sometimes called "default mode" activity, involve reduced prefrontal executive control and increased spontaneous activity in memory systems. The phonological loop runs more freely in the absence of competing cognitive demands.

Boredom: Explicitly boring or repetitive tasks show elevated earworm rates. The brain, under-stimulated, may activate musical imagery as a form of endogenous stimulation.

A study by Williamson and colleagues tracked earworm triggers in diary entries from over 300 participants. The most common initiating event was simply hearing the song (either in person or on media). The second most common was an internal semantic trigger — thinking of or discussing something with a lyrical association. Less commonly, earworms appeared entirely de novo with no identifiable trigger, suggesting that some activations are truly spontaneous retrievals from long-term memory.

The trigger often explains why certain songs become earworms at certain moments: a song associated with a particular period of your life may be unconsciously activated by similar contextual cues — a season, a time of day, a smell — that were present during that period. This is memory associativity at work.

Trigger Type	Mechanism	Susceptibility
Recent exposure	Fresh semantic activation, high availability	Very high
Semantic/word association	Lyrical or conceptual link in conversation	High
Stress or cognitive load	Executive system depleted, automatic processes dominant	Moderate to high
Idle/routine activity	Low executive demand, default-mode activation	Moderate
Boredom	Under-stimulation, endogenous activation seeking	Moderate
Emotional significance	Long-term memory strength via amygdala encoding	Moderate
Visual/auditory cue	Implicit associative activation	Low to moderate

Source: Williamson, V.J., Müllensiefen, D., & Stewart, L. (2014), "Earworms: Frequency of Involuntary Musical Imagery," Psychology of Popular Media Culture, 3(4), 231–240 — large diary study of earworm triggers and context.

Why Earworms Loop Rather Than Play Through

A fundamental puzzle about earworms is why they repeat a fragment rather than generating a complete playback of the song. When you deliberately recall a song, you can often reproduce it more or less intact, beginning to end. But earworms almost never do this. They lock onto a 15–30 second fragment and cycle it.

One explanation involves the distinction between conscious retrieval and automatic rehearsal. Deliberate recall involves goal-directed search through long-term memory: "Play me this song from beginning to end." The prefrontal cortex orchestrates the search, and sufficient executive resources allow a complete traversal. Involuntary imagery, by contrast, lacks this executive scaffolding. It runs in working memory, where capacity is limited and attention is not directed toward completion.

A second factor is the unresolved expectation hypothesis, mentioned earlier. The melody contains a structural feature — an unexpected interval or harmonic tension — that violates the listener's internalized musical expectations. The brain, attempting to resolve this through anticipatory predictive coding, re-generates the fragment hoping for resolution. Each cycle through the loop re-encounters the same violation, and the unresolved expectation prompts another cycle. The earworm becomes a cognitive itch that rehearsal attempts to scratch but cannot quite reach.

A third mechanism involves what researchers call "involuntary search processes." Once a melody enters working memory, the brain's implicit retrieval systems attempt to match it against stored patterns. If the match is incomplete (the pattern is not fully resolved, or there are multiple partial matches), the search continues in the background. This ongoing search manifests as continued cycling of the melody.

Importantly, earworms also eventually decay. The longer an earworm persists without fresh activation, the more the melody fades from working memory through the natural decay of the phonological store. This is why deliberate engagement with other material, or sleep, typically resolves even persistent earworms.

Source: Liikkanen, L.A. (2012), "Musical Fantasies — Mechanisms and Functions of Spontaneous Musical Imagery," Frontiers in Psychology, 3, 92 — theoretical framework distinguishing involuntary imagery from conscious recall and explaining persistence.

Getting Rid of an Earworm: Cognitive Mechanisms of Suppression and Displacement

If earworms arise from the phonological loop's automatic rehearsal, can conscious strategies actually disrupt them? The answer is yes, though not through willpower alone.

Engaging working memory: The most evidence-backed technique is to occupy the phonological loop with competing material. Chewing gum is the most famous example; multiple studies have found that chewing gum reduces earworm intensity and duration. The mechanism is thought to involve two processes: first, the motor and auditory components of chewing engage the articulatory rehearsal system (the system that rehearses the melody), so the system is occupied with processing the gum chewing and cannot simultaneously rehearse the melody. Second, gum chewing may trigger proprioceptive feedback (awareness of movement) that competes for attention. The effect is modest but reliable.

Other working memory engagement techniques include: solving a puzzle or arithmetic problem, playing a cognitively demanding game, engaging in conversation, reading challenging prose, or performing any task that demands auditory or verbal working memory. The specificity matters — the engaged task must use overlapping neural resources with the earworm generation. Purely visual tasks (playing a video game with no speech component) are less effective.

Completing the song: Counterintuitively, deliberately completing the earworm by mentally playing through the rest of the song often resolves it. If the unresolved expectation hypothesis is correct, this completion satisfies the brain's prediction system and allows the loop to release. Anecdotally, many people report that deliberately listening to the full song or playing it through mentally terminates the earworm. This is the inverse of the "don't think about it" approach — sometimes the antidote to an itch is to scratch it.

Replacing with a competing melody: Deliberately retrieving a different song (one that is more complex or less rehearsal-prone) can sometimes displace the earworm. This is essentially competing activation: the new melody occupies working memory and reduces resources available for the original earworm. However, this technique has a risk: the new melody may itself become an earworm.

Cognitive suppression (limited effectiveness): Deliberate attempts to "not think about the song" have mixed results. The irony is that suppression attempts often activate the memory representation more strongly — the phenomenon of rebound effects in thought suppression. Trying harder not to think of the earworm may strengthen its activation.

Distraction and time: The simplest approach is to engage in absorbing activity (work, conversation, exercise) that captures attention, and allow the earworm to decay naturally. Most earworms disappear within hours without intervention.

No technique works universally for all people or all earworms. Individual differences in musical imagery vividness, working memory capacity, and attention control all moderate the effectiveness of these strategies.

Source: Beaman, C.P. (2016), "Gum Chewing, Acoustic Distraction, and Spontaneous Imagery Suppressibility in a Cross-Linguistic Sample," Cognitive Processing, 17(4), 429–436 — experimental evidence that gum chewing suppresses involuntary musical imagery via phonological loop occupation.

When Earworms Are Intrusive: Pathological Musical Imagery and Related Conditions

For most people, earworms are mildly annoying but transient. But for a small subset, involuntary musical imagery becomes clinically distressing — persistent, intense, difficult to suppress, and interfering with sleep or concentration.

Severe involuntary musical imagery can occur in association with obsessive-compulsive disorder (OCD). Individuals with OCD report intrusive melodies that persist despite distressing and repeated suppression attempts, often accompanied by a sense of compulsion to mentally "complete" the song or resolve it in a specific way. The music itself becomes an obsession — the person is bothered not merely by the intrusion but by what the intrusion might mean about their mind or identity.

Involuntary musical imagery has also been reported in association with tinnitus (persistent ringing in the ears), where auditory hyperactivity may increase susceptibility to spontaneous auditory imagery. Some individuals with musical training or absolute pitch (perfect pitch) report higher rates of involuntary musical imagery, possibly because their stronger auditory-musical representations in long-term memory are more readily activated.

In rare cases, extremely vivid and persistent involuntary musical imagery has been reported following epileptic seizures, brain injury, or stroke affecting auditory regions, suggesting that certain forms of involuntary imagery can reflect genuine pathological activation rather than a normal exaggeration of the phonological loop phenomenon.

For those experiencing severe, distressing involuntary musical imagery, cognitive-behavioral techniques (related to OCD and intrusive thought management) and, in some cases, mindfulness-based approaches have shown benefit. The key distinction is that normal earworms are a benign artifact of memory and attention; pathological imagery involves subjective distress and functional impairment.

Source: Hemsley, D.R. & Margerison, J.H. (1988), "A Study of Some Aspects of Musical Hallucinations in Patients with Acquired Deafness," Journal of Psychosomatic Research, 32(3), 313–323 — early clinical documentation of involuntary musical imagery in sensory-deprived populations and its relationship to musical expertise.

Individual Differences: Why Some People Get Earworms More Than Others

Not all people are equally susceptible to earworms. About 10% of people report never or very rarely experiencing involuntary musical imagery; others report it daily. What accounts for these individual differences?

Musical training and expertise: Musicians and people with formal musical training report higher rates of involuntary musical imagery. This likely reflects stronger representations of musical material in long-term memory and greater automatic activation of auditory imagery during musical listening. However, the relationship is not linear — extreme expertise does not necessarily increase earworm susceptibility indefinitely.

Personality: Openness to experience (a personality dimension associated with creativity, curiosity, and engagement with aesthetic experience) correlates positively with earworm frequency. This may reflect a general tendency toward vivid mental imagery across domains.

Working memory capacity: Individuals with higher verbal working memory capacity actually show higher rates of involuntary musical imagery, contrary to initial intuitions. One explanation is that higher capacity allows more elaborate and persistent activation of auditory representations.

Attention control: Paradoxically, better attentional control does not necessarily mean fewer earworms. Instead, attentional capacity and the ability to flexibly engage or disengage attention may be more predictive. People who are good at directed attention but less adept at flexible disengagement may experience more persistent earworms.

Auditory sensitivity: Some people report more frequent and vivid involuntary musical imagery and also report heightened general sensitivity to sound. This could reflect individual differences in auditory cortex responsivity or in the signal-to-noise ratio of auditory processing.

Depression and anxiety: Elevated levels of anxiety and depression are weakly associated with increased earworm frequency, particularly intrusive, distressing earworms. This may reflect both the attentional effects of mood disorder (rumination, mind-wandering) and the tendency for negative mood to enhance the salience of intrusive thoughts or imagery.

These individual differences suggest that earworms are not purely a memory phenomenon but emerge from interactions between memory, attention, personality, and emotional state.

Source: Liikkanen, L.A., Jakubowski, K., Weird, F., & Thompson, W.F. (2016), "Affective Responses to Everyday Sounds and Myalgic Encephalomyelitis/Chronic Fatigue Syndrome," Frontiers in Psychology, 7, 1915 — population survey linking personality and cognitive factors to involuntary musical imagery frequency.

The Takeaway

An earworm is not a defect in musical taste or memory; it is a normal artifact of how the brain maintains acoustic information in working memory and how the auditory prediction system processes melody. The phonological loop, evolved for language, seamlessly applies itself to music — keeping melodies "alive" through automatic rehearsal. Melodies most prone to earworm status are those that are simple enough to generate effortlessly but structured in a way that leaves an expectation unresolved, prompting repeated cycling attempts. Triggers are straightforward: recent exposure, stress, idle mental states, and boredom all increase susceptibility. When an earworm does lodge itself, engagement of competing working memory tasks (gum chewing, problem-solving), deliberate completion of the song, or simple time and distraction are effective countermeasures. For the vast majority of people, earworms are transient curiosities that reveal how memory, prediction, and attention collaborate in consciousness. They are neither pathological nor evidence of musical obsession, but rather a window into the automaticity and vivacity of the mind's auditory imagery systems.

Frequently Asked Questions

Q: Why do catchy songs get stuck in your head more than complex music?
A: Catchy songs have a combination of features that make them ideal for involuntary replay: simple, predictable melodic contours that are easy for the brain to generate and maintain in working memory, regular rhythm that provides a stable scaffold, and often a single unexpected interval that creates an unresolved expectation. This combination maximizes both the ease of automatic rehearsal (simplicity) and the drive to continue rehearsal (incompleteness). Complex, unpredictable music provides fewer predictable patterns for the phonological loop to lock onto, so it rarely becomes an earworm.

Q: Is it bad to have an earworm stuck in your head?
A: For most people, earworms are harmless and benign — a minor annoyance lasting minutes to hours that resolves on its own. They do not indicate pathology or obsession. However, for a small subset of people, particularly those with OCD or heightened anxiety, involuntary musical imagery can become distressing and persistent, interfering with sleep or concentration. In these cases, cognitive-behavioral strategies or mindfulness approaches may help. The distinction is whether the earworm is mildly annoying or genuinely interferes with functioning and causes subjective distress.

Q: Does chewing gum actually remove an earworm?
A: Yes, with modest but reliable effectiveness. Chewing gum engages the same articulatory rehearsal system (part of the phonological loop) that maintains the melody, so gum chewing competes for neural resources and reduces the intensity and duration of the earworm. Multiple studies have demonstrated this effect. The mechanism appears to involve occupation of the motor and auditory components of the rehearsal system, and possibly proprioceptive feedback from jaw movement. The effect is not universal — it works better for some people than others and may depend on the characteristics of the earworm — but it is one of the few techniques with consistent empirical support.

Q: What is the difference between remembering a song and having an earworm?
A: Remembering a song (voluntary musical imagery) is a conscious, goal-directed retrieval where you deliberately decide to recall a song and can often reproduce it more or less completely, from beginning to end. An earworm (involuntary musical imagery) initiates spontaneously without intention, typically locks onto a fragment rather than the complete song, and persists despite conscious attempts to suppress it. Earworms feel automatic and are experienced as intrusions, while voluntary recall feels volitional. The neural difference involves the degree of executive control and prefrontal involvement: voluntary imagery recruits more dorsolateral prefrontal regions that coordinate goal-directed retrieval, while involuntary imagery arises more from automatic memory and attention systems.

Q: Why do I get earworms when I'm stressed or tired?
A: Stress and fatigue deplete the cognitive resources (executive function, attention control) that normally keep involuntary processes in check. When you are tired or cognitively overloaded, your brain shifts toward more automatic, less controlled processing. The phonological loop runs more freely without executive suppression, and memory representations that are emotionally salient or recently activated become more likely to spontaneously activate. Additionally, idle mind-wandering — common during stress and fatigue — provides the mental state (low attention, low prefrontal control) in which earworms are most likely to emerge. Once an earworm starts, fatigue also means less capacity to actively suppress or displace it.

Q: Is an earworm the same as hearing voices?
A: No. An earworm is involuntary musical imagery — the internal experience of a melody or song in the absence of external sound. The person knows it is not real sound; it is a memory phenomenon. Auditory hallucinations (hearing voices) involve the false belief or at least the impression that the sound is externally generated, not internally generated. Hallucinations can occur in conditions like schizophrenia, bipolar disorder, or delirium, and involve genuine disruption of reality contact. Earworms are normal, the person retains full reality contact (they know the music is internal), and approximately 90% of people experience them. They are not symptoms of mental illness but rather normal variations in how memory and attention operate.

---

Read next:

• How Music Affects the Brain
• Why Music Affects Emotions: The Science
• The Science of Learning: How We Learn