The first study into exactly what makes a song catchy has found that fast songs that put an unexpected twist on generic pop melodies are the most likely to get stuck in your head. The researchers have even come up with algorithms that can predict which tunes will be catchy and which won't.
Catchy songs – sometimes know as earworms – often have a generic rising and falling melody, with a few variations to make them distinctive and memorable, study author Kelly Jakubowski of Durham University told IBTimes UK.
The features that give a song its sticking power are a fast pace coupled with a melody shared with other popular tunes – often rising and then falling throughout a musical phrase – with an element of the unexpected thrown in. These can be an unusually big leap in pitch or using an amount of repetition uncommon in pop songs.
"Earworm tunes tend to follow these quite common melodic shapes that you often see in nursery rhymes, but that you also see in the chorus of Bad Romance by Lady Gaga, for instance, where the melody starts quite low in pitch, it rises in stepwise motion and then it falls again," Jakubowski says.
"I think of it as the brain searching for the optimal level of complexity in a melody. So something that's fairly simple but not too simple. It has to have some sort of level of interest to it."
The study confirmed that songs that do well in the charts and get more time on the radio are more likely to be classed as earworms.
Reverse engineering a catchy tuneJakubowski and her colleagues surveyed 3,000 people to find out what their most troublesome earworms were. They took these top earworm tunes and compared them with a list of tunes that had done equally well in the charts but weren't named by any survey respondents as an earworm.
The team then used a computer to analyse 83 different features of the tunes, such as the pitch range of the songs, how fast they were and how rhythmically diverse they were. They then used algorithms to predict whether a song was an earworm or a non-earworm based on their features they'd analysed.
What about the words?
Analysing how the lyrics of a song affect its catchiness wasn't a part of Jakubowski's study, but it's something she's planning to consider in the future. "That's something that nobody's looked at yet. We'd really like to look at the rhymes in the song or alliteration and other literary elements to see how they add to the 'earworminess' of a song as well."
As well as studying earworms, Jakubowski says that she frequently suffers from them. "I am a musician as well, so sometimes it's the music that I'm playing. But right now, because I keep talking about Bad Romance by Lady Gaga, I haven't heard the song in ages, but I've had it in my head constantly for the past few days."
The computer was then able to work out which songs of the 200 songs were earworms and which weren't. "The next step would be to test these algorithms so we could put in a new song and predict whether it will be an earworm," she says.
Jakubowski envisions that this could be used by songwriters or producers in the future to help them try to make a hit tune. "A songwriter might write a song and you'd put it into the algorithm and try to get a catchiness measure."
The study is published in the journal Psychology of Aesthetics, Creativity and the Arts.