Science Agrees: Pop Music Really Does Sound the Same


When I was younger, I was convinced that the pop music of the future would involve banging on metal garbage can lids. To a kid who grew up on the Beatles, Mariah Carey and Boyz II Men were just about as different as music could possibly get. Surely the only uncharted territory that remained was garbage can lids.

I'm not aware of any popular genre of music involving garbage cans, so I must admit I was wrong with that prediction. I assumed that popular music would continue to evolve unabated, but it turns out that this is not the case. Even science says that all pop music really does sound the same.

Using computers to analyze the so-called "Million Song Dataset," researchers at the Spanish National Research Council sifted through 464,411 different recordings to understand whether music had changed in any definable way between 1955 to 2010 (their study was published in Scientific Reports last month).

To use tools from the study of statistical physics and complex networks, the music had to first be converted into something that could be quantified. To do this, the researchers transformed three different dimensions of music into "codewords."

For example, one dimension they examined was pitch, which describes things like chords and melody. There are 12 possible values for pitch (C, C#, D, D#, etc.), and each part of a song that is analyzed will get a yes or no answer for each of those 12 values. That means that there are 2^12, or 4,096, possible pitch codewords. The number of times each of these 4,096 codewords appeared in a sample was counted, and then fit mathematically. What this showed was that a few codewords showed up very frequently, but most of the codewords were very rare. In other words, most music was very similar. Interestingly, these patterns were highly stable -- the frequencies of pitch distribution remained constant over more than 50 years of music.

Of course, just because the same codewords are being used at the same frequencies doesn't mean that the songs are the same. Just as with language, similar words strung together in different ways can have completely different meanings. To quantify the syntax of the codewords, they measured transitions between codewords and mapped out these connections.

By measuring the number of links each codeword has to other codewords, they found that the "small-worldness" of the networks had decreased over the years. In recordings from the more recent years, it took more steps to get from one codeword to another, and each codeword was less connected to the rest of the network. In other words, the transitions between pitches are more defined and limited with the more recent music.

They also saw less variety over the years in timbre, the dimension of music associated with things like different types of instruments. Starting with 177,147 possible timbre codewords, they saw that more frequent codewords became even more frequent, and the rare ones became even more rare.

And finally, just as your parents have been saying for years, music really is getting louder. As expected, loudness has increased over time, but the range of loudness values within each recording has remained constant. So even though music has gotten louder over the years, the dynamic variability within a track is being preserved. While that's good news for now, digital media can only get so loud (0 dBFS), so if this loudness war keeps going then dynamic variability will eventually be restricted.

Even though the study found that music is getting louder and more homogenized over time, the similarities between the different recordings also remind us that pop music is essentially pop music whether recorded in 1955 or 2010. Maybe this inherent stability of pop music can explain why the Beatles Anthology was able to capture the interest of the popular kids in my sixth grade class, who all of a sudden decided that my taste in music wasn't actually so strange.

Photo credit: VARA/Wikimedia Commons




Add new comment

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd> <p> <div> <br> <sup> <sub>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.