Mapping emotions in bad religion’s music : insights from qualitative and automatic sentiment analysis

In this chapter, we analysed Bad Religion’s seventeen studio albums’ emotional expressions using automatic sentiment analysis to map the emotional trends over time, and to evaluate how these emotions are manifested. The automatic sentiment analysis revealed prevailing negative or neutral tones, depending on the model used (i.e. the Ekman model and the Go Emotions model). The most prominent emotions also vary, from fear, sadness, and anger, on the one side, and annoyance, and disappointment, on the other. Despite slight variations across the albums, no distinct upward or downward trends were identified for specific emotions. The qualitative analysis disclosed that most emotion-related terms occur infrequently (one - nine occurrences) in the lyrics. We assessed the use of emotion words in expressing the singer persona’s emotions or addressing emotions more broadly, revealing four patterns: authenticity, inclusion, external reality, and storytelling. Lastly, our topic analysis on words associated with emotional triggers supported previous findings, namely that conflict, justice, mind, and belief are common themes in Bad Religion’s lyrics, often expressed with negative terms.