Sad songs say so much: where emotion meets repetition in pop music

On the closing track of his hit album Breaking Hearts, Sir Elton John made the observation that “Sad Songs (Say So Much);” and it is as true today as it was back in 1984.

In a 2017 article for The Pudding, Colin Morris took a look at how music from the 1960s to the 2010s gets more repetitive decade after decade. But the old Elton John song got me thinking: can you also find a correlation between emotion and a song’s repetitiveness? The data comes from taking song lyrics — just the words — from Billboard’s All-Time Hot 100 list and running them through a simple data compression tool.

Before we get there, let’s start with the standards. When I think of a sad song, the first thing that pops into my head is the Adele hit “Someone Like You.” The compression tool can store the lyrics to that song with 43% of the space necessary to store the same number of uncompressed text characters. This seems like a lot of compression, but keep in mind that it does not only deal with repeated words. The algorithm also captures the same repeated letter patterns in two different words. In contrast, the song “Happy” by Pharrell Williams can be stored with 26% of the space. It is much more compressed and therefore much more repetitive.

You could say the Adele song is able to be compressed 57%, and the Pharrell song is able to be compressed 74%. Ninety-nine of 100 songs on the Billboard list have lyrics. Number 25 is “The Theme from ‘A Summer Place,’” which happens to be an instrumental.

Pop songs run the gamut of emotions, beyond just happy and sad. They can be calming or angry. There can also be a combination of emotions in a single song. For the purposes of this experiment, I’m only looking to see if a song is generally more similar to the Adele song above or the Pharrell song above. I will call it the Adele/Pharrell scale. By this scale, 40 of the 99 songs are generally sad and 59 of the songs are generally happy. In the series of graphics below, the low-compression outliers tend to be Adele-like songs; and the Pharell-like songs tend to be the high-compression outliers.

It makes sense when you think about it. Artists who are inspired by a loss bare their soul in their lyrics. Artists who convey happiness are only required to find a catchy hook and repeat it to a melody that makes listeners feel like dancing.

-The article which inspired this can be found here: https://pudding.cool/2017/05/song-repetition/

-The compression tool used in this article can be found here: http://www.unit-conversion.info/texttools/compress/

-The Billboard All-Time Hot 100 list can be found here: https://www.billboard.com/charts/greatest-hot-100-singles/

-The database I created by compressing lyrics can be found here: https://docs.google.com/spreadsheets/d/1_H29yr5HzNsOY0iSatBbGSTyfvqcNlBbgeHqON17W1I/edit?usp=sharing