Why Do Songs Sound the Same: Information Cascades in Popular Music
https://www.nytimes.com/interactive/2018/08/09/opinion/do-songs-of-the-summer-sound-the-same.html
https://thenextweb.com/artificial-intelligence/2017/10/19/ai-music-hit-song-predict/
A rather annoying phenomenon in radio, especially in Top-40 radio, is stations playing the same few popular songs over and over again. For example, it is not uncommon to be listening to a radio station, hear a specific song, switch to a different station, and hear that new station playing the exact same song.
Another example of homogeneity in popular music would be overall song structure in general. An interactive New York Times article called “Why Songs of the Summer Sound the Same” details how “Songs of the Summer” usually have similar structures, and that popular songs tend to sound similar to each other. For example, in the summer of 2010, “California Girls” by Katy Perry, “Your Love is My Drug” by Kesha, and “Alejandro” by Lady Gaga, as well as 7 other top songs of the summer all had similar acoustic characteristics, or musical “fingerprints.” Now, there’s even an entire branch of recent research called “Hit Predicting” which uses AI to try to predict which songs will become popular based off factors such as valence, tempo, and and key. So how did these “musical fingerprints” get so popular?
The answer, like many trends in fashion, food, technology and more, lies in “herding” or “information cascading.” An information cascade has the potential to occur when (1) people make decisions sequentially, (2) later people watch the actions of earlier people, (3) those people infer something based off what they observed other people doing, (4) but not the private information those people possess. An information cascade actually occurs once people abandon their own information and instead make decisions based off earlier people’s actions. In other words, they make decisions based off what they see the crowd doing, but not based off knowing what the crowd knows.
Furthermore, marketers can use the notion of information cascading to popularize certain items. Let’s take a look at how this example would work with a popular song from the New York Times article. The article points out that Top-10 songs today are more homogenous than they were in, say, the 80’s because there are fewer popular genres and a small number of songwriters make up a large percentage of the top hits. Most notably is Swedish producer Max Martin, who is credited with an astonishing 22 #1 singles.
Let’s apply the popularity of the “Max Martin” sound to Information Cascading. Say, for example, you are a rising songwriter who wants to make a hit record. Using your own personal research, you decide that you’d like Rick Rubin to produce your new album. However, it seems like every time you turn on the radio you hear another song produced by Max Martin, and lots of your peers in the music industry use him as a producer. Despite not knowing how those artists feel about their experiences recording with Max Martin (you haven’t spoken to any of them about it), you assume that with tons of famous clients and songs he’s doing something right, so you decide to have him produce your record instead.
If we set this case up as an information cascade, the songwriter is the person with personal, private information. (1) He makes his decision for finding a producers after countless musicians before him already have, (2) since he is in the music industry, he watches the actions of earlier musicians, (3) he infers that people are satisfied with Max Martin because so many people have used him as a producer, and (4) he does not know for sure whether Max Martin’s clients are actually satisfied with their experience recording an album with him. Thus, our songwriter ignores his own private information and instead uses the “wisdom of the crowd” to record an album with him. Once this trend “cascades,” or catches on, Max Martin rises in popularity, his “musical fingerprint” dominates Top-40 radio, and if we apply this logic to overall musical trends, “songs of the summer” sound the same!