The Song That Won't Leave

Somewhere in the last few minutes, a song you did not choose may have started playing inside your skull. Not the whole thing. A fragment: a hook, a chorus, four words and a tune that loop and loop and refuse to resolve. You did not press play. You cannot press stop. And the more you notice it, the more insistent it becomes.

This is one of the most common experiences in human cognition, and for most of history nobody bothered to study it. In surveys, roughly 98 percent of people report having had a song stuck in their head ¹. Almost no one escapes. It happens to children and pensioners, to people who love music and people who claim not to care about it at all. And yet for centuries it sat below the threshold of scientific curiosity, dismissed as a triviality, a quirk too small to dignify with research.

That changed only recently. The phenomenon now has a clinical name, several competing theories, and a growing body of experimental data. The name itself was borrowed, like so many words for things English never quite captured, from German. Ohrwurm: ear-worm. A fragment of music that crawls into the mind and burrows in. The image is grotesque and exactly right.

A modern affliction

For almost all of human history, music behaved differently than it does now. It was an event, not an object. When the musicians stopped playing, the music stopped existing. You could carry a tune in your memory, hum it, teach it to someone else, but you could not summon the original performance on demand. Repetition was rare. To hear the same song twice you had to make it happen, deliberately, through effort.

The twentieth century dismantled all of that. Recording, then radio, then the cassette and the compact disc and finally the streaming feed turned music into something that could be repeated infinitely and identically. The same three-minute song could play in a million kitchens at once and play again the next morning unchanged. Advertising jingles were engineered to lodge in memory. Pop choruses were refined, decade by decade, into ever-stickier shapes. The air filled with short, catchy, endlessly repeated musical fragments.

It is no coincidence that the scientific study of earworms is a creature of this saturated environment. The man often credited with starting it was not a neuroscientist but a marketing professor. James Kellaris, who taught at the University of Cincinnati, became fascinated in the late 1990s by the songs that seemed to haunt his students. He surveyed hundreds of them and found the now-famous figure: nearly all of them, around 98 percent, reported the experience ¹.

Kellaris reached for a metaphor that has stuck almost as stubbornly as the songs themselves. He called the earworm a cognitive itch. The idea is intuitive. An itch is an irritation that demands attention, and the act of scratching provides momentary relief while doing nothing to remove the cause. In Kellaris’s framing, the mind “scratches” a stuck song by replaying it, which satisfies the urge for an instant and then leaves the itch exactly where it was. The loop perpetuates itself.

The metaphor was vivid, but it left the central question untouched. If almost everyone gets earworms, why do only certain songs become them? Why does one chorus drill into your head for a week while a hundred others wash past and vanish? To answer that, someone had to stop asking people whether they got earworms and start examining what the earworms actually were.

The anatomy of a sticky melody

That someone was Kelly Jakubowski, a researcher who, working at Durham University in England, decided to treat earworms as musical objects worth dissecting. Instead of relying on impressions, she gathered a large dataset of songs that people reported getting stuck in their heads and compared them, systematically, against songs that were equally popular but never seemed to lodge. If earworms shared a hidden structure, the comparison should reveal it.

Her 2016 study, published in the journal Psychology of Aesthetics, Creativity, and the Arts, found exactly that ². The sticky songs were not random. They clustered around a particular musical profile, and that profile turned out to be surprisingly specific.

The first feature was tempo. Earworms tended to be faster, up-tempo songs, the kind with a pace that invites you to hum or tap along. This makes a certain sense. A brisk melody is easy to carry, easy to reproduce internally, easy to keep spinning.

The second feature was shape, what musicians call contour: the way a melody rises and falls. Earworms tended to follow a common, intuitive contour, the sort of up-and-down arc that feels natural to sing. Think of the opening of a nursery rhyme or a stadium chant. The melody behaves the way a human voice wants to behave, which makes it easy to learn and easy to recall.

But a song that is only easy and familiar would be forgettable in a different way. The boring stuff fades. So Jakubowski found a third, more interesting feature, almost a paradox. Earworms hid something inside the familiar contour: an unusual interval, an unexpected leap or pause, a small moment that violated the pattern the listener was expecting. The melody was conventional enough to learn at once and strange enough to refuse to settle.

This is the crucial insight. The most effective earworms are not simply the catchiest songs. They are catchy and a little surprising. They lure the brain in with a familiar shape and then snag it on an unresolved oddity. The mind keeps returning to the spot where its prediction failed, trying again to make it fit. The very thing that makes the melody memorable is the thing that keeps it from feeling finished.

The pull of the unfinished

To understand why an unresolved fragment should loop, it helps to go back nearly a century, to a Lithuanian-born psychologist named Bluma Zeigarnik. The story of her discovery has the quality of a parable. Sitting in a Vienna cafe in the 1920s, she is said to have noticed that the waiters could remember complicated, unpaid orders with ease, yet seemed to forget them entirely the moment the bill was settled. The completed transaction vanished from memory. The open one stayed sharp.

Zeigarnik turned the observation into experiments. She gave people tasks to perform, interrupting some before completion and letting others finish. Afterward, she asked which tasks they remembered. The interrupted ones came back far more readily than the completed ones ³. The finding became known as the Zeigarnik effect: unfinished tasks occupy the mind more persistently than finished ones. An open loop holds a kind of cognitive tension that a closed loop releases.

It is not hard to see how this maps onto music. A melody, especially a tonal one, sets up expectations. It implies a direction, a resolution, a place where it ought to come to rest. When you have only a fragment, a chorus without its verse, a phrase that loops back before reaching its cadence, the song behaves like an interrupted task. Your mind treats the missing ending as a problem to be solved, and it tries to solve it the only way it can, by playing the loop again, hoping this time to reach the resolution that never comes.

The numbers support the picture. Earworms tend to consist of a short segment, lasting on average around twenty seconds before looping back to the start ¹². That brief, repeating fragment is almost never the whole song. It is the part the brain cannot resolve, the open loop it keeps reopening. The completed portions of the song, the parts that reach their natural close, do not haunt you. Only the unfinished sliver does.

When the loop slips in

If earworms are unresolved fragments waiting for a chance to surface, the obvious question is when that chance arrives. The answer, consistently, is when the mind has room.

Earworms strike hardest in low-attention states, the moments when consciousness is idling and attention is not pinned to anything demanding. The shower is the classic example. So is walking, doing the dishes, driving a familiar route, lying awake before sleep. In these gaps, the mind wanders, and a stray melodic fragment can rise into the vacancy and start to spin. A demanding task, by contrast, tends to crowd earworms out; there is no spare bandwidth for them to colonize.

Stress and fatigue widen the door further. A tired or anxious mind is a more porous one, less able to hold its own focus, more prone to intrusive loops of all kinds. Earworms travel in the same company as rumination and other unbidden thoughts. They tend to arrive when the gatekeeping faculties are worn thin.

Exposure matters too. The Finnish researcher Lassi Liikkanen, who studied earworms in large surveys, found that they were strongly linked to recent and frequent musical exposure ⁴. Hear a song often enough, or recently enough, and you effectively stock the loop with ammunition. The fragment is primed, sitting just beneath the surface, ready to surface the moment your attention drifts. This is why the song from the radio that morning, or the one you played four times on a long drive, is the one that ambushes you that night.

The frequency of all this is striking. In Liikkanen’s work, around 90 percent of people reported getting earworms at least once a week ⁴. Musicians and heavy listeners, the people whose minds are most thoroughly furnished with music, reported them even more often. The more music you carry, the more material there is to loop.

Not a bug, but a feature

Here the story takes a turn that overturns the whole framing. We tend to describe an earworm as a malfunction, a glitch, the brain caught in a faulty loop it cannot break. But there is a growing case that this gets the situation almost backwards.

Consider what an earworm actually does. It takes a fragment of music and rehearses it, repeatedly, in the absence of any external input. From the standpoint of memory, that is not an error. That is consolidation. Repetition is how the brain strengthens and stores patterns, and a melody that loops on its own is a melody being practiced. Some researchers now suspect that earworms are, at least in part, a byproduct of the ordinary machinery of musical memory: the same processes that let you recall a tune at all, running on idle, keeping the pattern alive.

Seen this way, the annoying song is not a system failure. It is a learning tool that happens to have surfaced into consciousness. Your brain is doing something it is very good at, the kind of thing that, over evolutionary time, would have helped a mind hold onto songs, stories, rhythms, and the long chains of information that oral cultures preserved entirely in memory. We notice earworms only when they intrude on attention. But the underlying rehearsal may be running far more often than we realize, quietly maintaining the vast library of music we carry without effort.

This reframing does not make the experience less irritating in the moment. But it does change what the irritation means. The stuck song is not evidence that something has gone wrong. It is evidence of how thoroughly musical the human mind is, how readily it absorbs melody and how diligently it works to keep what it has absorbed.

How to make it stop

None of which helps much when a four-word loop has been spinning since breakfast. So the practical question remains: how do you evict a stubborn earworm?

The research offers a few genuine answers, and one strong warning. The warning first: do not fight it directly. Trying to suppress an unwanted thought tends to backfire, a phenomenon documented in the psychology of thought suppression. Tell yourself not to think of something and you summon it. The harder you push against the loop, the louder it tends to play. Resistance feeds it.

The more effective strategies work by redirecting rather than wrestling. One of the most surprising comes from a 2015 study at the University of Reading, which found that chewing gum measurably reduced both the frequency and the vividness of earworms ⁵. The proposed mechanism is mechanical. Inner singing recruits some of the same subvocal motor machinery used for actual speech and song, and the repetitive jaw movement of chewing seems to interfere with it. Occupy the apparatus of the inner voice with something else, and the song has less to sing with.

A second approach follows directly from the Zeigarnik logic. If the loop persists because it is unfinished, then finishing it may release the tension. Listening to the whole song, all the way through to its proper ending, can close the open task and let the fragment rest. The brain got the resolution it was looping for, so it stops looping.

A third strategy is plain distraction, but the engaging kind. A mildly absorbing task, a conversation, a puzzle, anything that occupies attention without being so dull that the mind wanders back to the music, can crowd the earworm out. The point is to fill the vacancy that let the loop in, rather than to attack the loop itself.

What unites the workable methods is that none of them is a fight. The cure for a stuck song is rarely effort and almost never anger. It is gum, completion, or gentle redirection, ways of giving the loop somewhere to go or something to do instead.

Coda

So the next time a song hijacks the quiet inside your head, on the walk home or in the shower or in the slow minutes before sleep, it may help to remember what is actually happening. A familiar melody snagged your attention on a small unresolved surprise. An unfinished fragment, behaving like Zeigarnik’s interrupted task, keeps reopening itself in search of an ending it cannot reach. And underneath the irritation, the machinery of musical memory is doing precisely what it evolved to do, rehearsing, predicting, holding on. The stuck song is not a malfunction to be fixed. It is the sound of a mind that is, whether you asked for it or not, profoundly musical.

Sources

1. Kellaris, J. A., “Identifying Properties of Tunes That Get Stuck in Your Head,” Proceedings of the Society for Consumer Psychology, 2001. — https://www.researchgate.net/publication/285672297_Identifying_properties_of_tunes_that_get_stuck_in_your_head
2. Jakubowski, K., Finkel, S., Stewart, L., Mullensiefen, D., “Dissecting an Earworm: Melodic Features and Song Popularity Predict Involuntary Musical Imagery,” Psychology of Aesthetics, Creativity, and the Arts, 2016. — https://psycnet.apa.org/record/2016-53727-001
3. Zeigarnik, B., “Das Behalten erledigter und unerledigter Handlungen” (On Finished and Unfinished Tasks), Psychologische Forschung, 1927. — https://en.wikipedia.org/wiki/Zeigarnik_effect
4. Liikkanen, L. A., “Music in Everymind: Commonality of Involuntary Musical Imagery,” Proceedings of the International Conference on Music Perception and Cognition, 2008. — https://www.researchgate.net/publication/228799821_Music_in_everymind_Commonality_of_involuntary_musical_imagery
5. Beaman, C. P., Powell, K., Rapley, E., “Want to Block Earworms from Conscious Awareness? B(u)y Gum!,” Quarterly Journal of Experimental Psychology, 2015. — https://journals.sagepub.com/doi/10.1080/17470218.2015.1034142

Related reading

• The Slot Machine in Your Pocket
• The Pleasure of the Safe Scream

More from the Mind edition →