Computer Analysis Shows Music Getting Angrier
Written by Kay Ewbank   
Saturday, 09 February 2019

A study of popular music by data scientists, using the IBM Watson Tone Analyzer, has found increasing levels of anger and sadness in the lyrics from the 50s till now.

Researchers Kathleen Napier and Lior Shamir from Lawrence Technological University in Michigan have published a research paper in the Journal of Popular Music Studies, showing how the lyrics from 6150 songs from the Billboard Hot 100 list have changed over the years. The team used quantitative analytics to study how the lyrics have changed. A text-mining program called Tone Analyzer, part of the computational linguistic tools in IBM Watson Developer Cloud. This service uses linguistic analysis to detect joy, fear, sadness, anger, analytical, confident and tentative tones found in text. There's a demo of Tone Analyzer that you can use to analyze text to see how it works, along with an API reference showing how to use Tone Analyzer in your own projects.



Tone Analyzer was used to apply automatic quantitative sentiment analysis, which associates all the words and phrases in a piece of text - in this case the song lyric - with a set of tones that they express. The individual tones are then combined to provide an overall sentiment. In the study, the sentiment measures for all the songs in the Billboard Hot 100 songs for a year were averaged, giving an annual score. This was then used to work out how levels for that sentiment - anger and sadness - changed over the decades.

The analysis showed that the hits released in the mid 1950s had the lowest levels of anger, though the three years from 1982-84 were equally low in angry lyrics. Generally, though, the lyrics became more angry and sadder over the analysis period.Levels of disgust and fear also grew over the years, though not as markedly as anger and sadness. Songs became more angry particularly in the mid 1990s, with levels peaking in 2015, according to the study. An equally depressing finding was that joy as a sentiment dropped steadily from the 1950s to present day, apart from a brief period in the mid 1970s. 


More Information

Quantitative Sentiment Analysis of Lyrics in Popular Music

IBM Watson Tone Analyzer

Tone Analyzer demo

Related Articles

IBM's Watson Gets Sensitive

A New Impetus For IBM Watson



To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on, Twitter, Facebook or Linkedin.


Microsoft Graph Data Connect Now Generally Available

Microsoft Graph data connect is now generally available. The announcement was made at this year's Build conference. Graph is a unified programmability model and APIs that can be used to access data wi [ ... ]

The End Of The App Store Era - Apple To Face Lawsuit

In October of last year we reported on Apple's appeal to the Supreme Court Of The United States to dismiss a lawsuit claiming that the App Store is a monopoly. SCOTUS has now granted permission for th [ ... ]

More News





or email your comment to:

Last Updated ( Saturday, 09 February 2019 )