Home / Blog

Why songs get stuck in your head

Published August 10, 2026

Songs get stuck in your head because a short musical phrase is cheap for your brain to replay and expensive to consciously switch off. Researchers call it involuntary musical imagery, and they have been measuring it for well over a decade.

The mechanism, as far as the evidence goes: recent exposure to a tune, or something that merely reminds you of one, primes a fragment of it. That fragment replays through the same inner rehearsal machinery you use for silent speech, which is part of why an earworm feels heard rather than thought about. Fragments are short and usually stop somewhere unresolved, so the loop restarts.

Four groups did most of the work below. Lassi Liikkanen surveyed 12,519 Finnish internet users and found 89.2% reported involuntary music at least weekly. Victoria Williamson and colleagues sorted real onset reports into four kinds of circumstance. Kelly Jakubowski's team compared earworm tunes against matched non-earworm tunes across 83 melodic features. Philip Beaman's group tested whether chewing gum shuts a loop down. It does, somewhat.

The short version

An earworm is a short melodic fragment replaying in your inner ear without being asked. It usually starts from recent exposure or a memory association, and it is close to universal. Occupying the machinery you speak with, gum included, reduces it.

Almost everyone gets them, and gets them often

Liikkanen put a number on it. Among 12,519 Finnish internet users, 89.2% said they experienced involuntary music at least once a week, and the more music people played and listened to, the more often it happened (Liikkanen, 2012). One country, one web-recruited sample, retrospective self-report, so treat it as an estimate rather than a constant of nature.

The wider picture is steadier. Liikkanen and Jakubowski later reviewed 47 studies drawn from 33 peer reviewed articles and called involuntary musical imagery a component of ordinary music cognition rather than a quirk, while noting the evidence leans heavily on Western, educated populations (Liikkanen and Jakubowski, 2020).

It is also not usually a torment. Across a thousand-plus open-text reports, many episodes were rated neutral or pleasant, with passive acceptance and plain enjoyment among the commonest reactions, though a real subset called theirs disturbing (Williamson et al., 2014).

The four ways an earworm starts

Williamson and colleagues gathered everyday onset reports from BBC Radio 6 Music listeners and an online survey, then classified them by grounded theory rather than by hypothesis. Four categories fell out (Williamson et al., 2012).

One detail is worth keeping: earworms were not always attached to overly familiar music, which is how a jingle you have heard twice and actively dislike can own your Tuesday. This was a qualitative classification of self-reports, so it maps circumstances rather than causes.

Some tunes are just built stickier

Jakubowski's group took the fun version of the question. Around 3,000 survey participants named the tunes that got stuck in their heads. The researchers pulled 100 frequently named earworm tunes and 100 tunes nobody named, matched them for popularity and style so familiarity alone could not explain the gap, then compared all 200 across 83 melodic features (Jakubowski et al., 2017).

The earworm tunes had more common global melodic contours, the overall rise-and-fall shape you find in a lot of nursery rhymes. They had less common average gradients between melodic turning points, a technical way of saying there is an unusual leap somewhere in the line. Their average tempi were faster.

A generic shape with one strange interval in it, delivered quickly. That is a surprisingly good description of most choruses you cannot get rid of.

The paper's title credits song popularity alongside melody, and popularity is the unglamorous half: heavy rotation still matters. What the matched comparison adds is that melody carries part of the load by itself. What it does not add is a recipe. Nobody has built one to spec.

How to get rid of one, gum included

The obvious move, deciding to stop thinking about it, is the one that reliably backfires.

People in the Sticky Tunes surveys reported better luck going the other way: engaging with the tune deliberately, hearing it through, or filling the same channel with other verbal or musical material. Some keep a designated cure tune. Plenty wait it out. These are self-reports about what felt effective, not controlled trials, so hold them loosely.

Then there is the gum, which is a genuinely good experiment wearing a silly hat.

Beaman, Powell and Rapley reasoned that if the loop runs through articulatory motor planning, the machinery you form speech with, then giving your jaw a job should interfere. Three experiments followed (Beaman, Powell and Rapley, 2015). Chewing gum reduced both deliberate and unwanted musical thoughts. The effect landed on the experience of hearing the music, not merely on thinking about a tune in the abstract. And it was not a generic attention or dual-task effect. The authors note this is the opposite of what thought-suppression theory predicts, since a concurrent task should make suppression harder.

So gum is not a cure. It is a wrench in the machinery, and a wrench working there is a decent clue about where the loop lives.

The flip side: a loop is rehearsal nobody paid for

Everything above treats the earworm as a nuisance to be managed. Language learners should look at it the other way around.

If the fragment looping in your head is in a language you are learning, the loop is review. Free review, on a bus, at a moment when nothing on earth could have talked you into opening a flashcard app. The words ride along because they are welded to the melody, and the melody is what your brain chose to replay.

That is why a song behaves differently from a word list. The memory research is in why music helps you remember vocabulary, and the honest evidence for songs as a study method is in do songs actually help you learn a language.

The limit is real. A loop repeats whatever it repeats and never asks whether the word is due. Replay makes a memory more frequent, not better timed, which is why a spaced repetition schedule still belongs underneath the music.

Get your first song free

What Lingotify does with the mechanism

Lingotify writes songs out of your own vocabulary list, so the lines that come back are made of words you actually need.

The piece built directly on this research is the Hook preset. It takes the words your review history says you are worst at, builds the chorus around them, and repeats that phrase at least four times across the song. Whichever fragment ends up looping afterwards, the odds are stacked toward the one carrying your weakest words. Williamson's most common trigger was repeated exposure; the Hook preset is that lever, aimed.

The rest follows the same logic. Karaoke lights up each word as it is sung. Quizzes appear inside the song and blank a vocabulary word out of the line you just heard. Every answer feeds an SM-2 spaced repetition schedule, the job no earworm will do for you.

Two songs a month are free, plus up to two more for watching an ad. Premium is $5.99 a month with a 7 day trial. For a language-specific walkthrough, start with learning German with music.

Make your first song free

Questions people ask

Why does only part of a song get stuck instead of the whole thing?

Because involuntary musical imagery is defined by fragments. Liikkanen and Jakubowski describe it as commonly consisting of the repetition of a short melodic fragment rather than a full piece. The part that loops is usually the part you have heard most, and it tends to end without resolving, which sends you back to the start of it instead of onward.

Does getting earworms mean something is wrong with me?

No. In Liikkanen's 2012 survey of 12,519 people, 89.2% reported involuntary music at least weekly, and the 2020 review of 47 studies frames it as a component of ordinary music cognition. Most episodes are rated neutral or pleasant, although a minority of people do find theirs genuinely disturbing.

Can you give yourself an earworm on purpose?

Partly. The most common circumstance in Williamson's classification was recent or repeated exposure, so playing one tune repeatedly is the closest thing to a deliberate lever, and melodic features seem to load the dice further. Nobody has demonstrated a reliable recipe. Lingotify's Hook preset works this angle by repeating a chorus built from your weakest words.

Sources