Pop music lyrics have become more rebellious and less empathetic over the past six decades

A recent analysis of popular music from the past six decades suggests that song lyrics have become increasingly focused on negative moral concepts like harm and rebellion. Over the same period, expressions of positive virtues like care and purity have declined. The findings provide evidence that mainstream music reflects broader societal shifts toward darker and more emotionally intense themes. This research was recently published in the journal Scientific Reports.

Music has always served as a mirror for culture, capturing the political, social, and emotional landscape of its time. Previous analyses of popular music indicate a rising trend toward individualism and self-promotion in modern lyrics. Studies also show a general decline in positive emotions like joy, alongside an increase in words expressing anger, fear, and sadness.

Lead author Vjosa Preniqi, a researcher at Queen Mary University of London, noted in a press release that this relationship between art and society inspired the project. “Music is much more than entertainment,” Preniqi said. “It is one of the ways societies tell stories about themselves. By analyzing song lyrics across several decades, we can begin to see how emotional expression and moral narratives evolve over time.”

To better understand these cultural shifts, researchers turned to a psychological framework known as Moral Foundations Theory. This framework proposes that human morality rests on five fundamental dimensions. Each dimension features a positive virtue and a corresponding negative vice.

The first dimension is care versus harm, which deals with empathy and cruelty. The second is fairness versus cheating, which covers justice and deceit. The third dimension is loyalty versus betrayal, relating to group cohesion and treason.

The fourth dimension is authority versus subversion, focusing on obedience and rebellion. The final dimension is purity versus degradation, which involves sanctity and corruption. By mapping these moral dimensions onto song lyrics, scientists can track how values evolve in popular culture.

Charalampos Saitis, an assistant professor of digital music processing at the Centre for Digital Music at Queen Mary University of London and lead of the Communication Acoustics Lab, explained why the team focused on these specific moral dimensions. “Music is one of the most widely shared forms of culture (music sociologist Tia DeNora has beautifully described music as ‘a technology of the self’), and as such it has long been studied as a barometer of cultural change, but such work usually relies on the musical audio content alone,” Saitis told PsyPost.

“When lyrics have been examined, this mostly has focused on mood or topic, and we wanted to ask a different question: what do songs express about our ideas and conceptions of right and wrong, and has that changed?” Saitis continued. “There is a growing body of research exploring moral expression in news articles, fiction books, and social media, but no previous work has considered popular song lyrics and how moral narratives can be constructed in popular entertainment media.”

To conduct their analysis, the researchers gathered text from two massive collections of song lyrics. The primary collection included 377,812 English language songs released between 1960 and 2010. This broad dataset featured music from 7,131 solo artists and 4,294 musical groups.

Because this first database lacked recent songs, the researchers compiled a second collection of 5,580 highly successful tracks from the Billboard Year-End charts. This secondary dataset covered the period from 1960 to 2023. This allowed the researchers to compare a massive general catalog against a narrower list of commercially successful hits.

To process the lyrics, the researchers utilized a specialized artificial intelligence model fine-tuned for text prediction. This computer program scanned the lyrics and generated a probability score for each of the ten moral polarities. A higher score indicated high confidence that the lyrics contained a specific moral theme.

The researchers also made this analytical tool available to the public. They created a web app based on the artificial intelligence language model used in the paper, known as the MoralBERTApp. Users can upload spreadsheet files containing text or lyrics, and the app will automatically assign probabilities for the ten moral categories.

The scientists did not just look at morality in isolation. They also examined the emotional tone and thematic content of the songs. They used established digital dictionaries that assign emotional values to thousands of words, allowing the team to measure feelings like joy, anticipation, and disgust.

Additionally, they used a statistical technique that groups words based on how frequently they appear together. This process identified broad themes within the text, which the researchers labeled with titles like violence and darkness or love and emotions. The team then looked for mathematical correlations between the moral scores and these emotional themes.

The analysis revealed a consistent shift in the moral tone of popular music over the last sixty years. Expressions of moral virtues, specifically care and purity, showed a gradual decline in both datasets. At the same time, lyrics associated with moral vices, such as harm, cheating, subversion, and degradation, experienced significant increases.

“What we found was a gradual shift away from language associated with virtues such as care and decency, towards themes that reflect conflict, harm and other moral concerns,” Preniqi said. “These patterns are dependent on various factors, such as genre and shock-factor, but they provide a fascinating window into changing cultural values and emotional expression.”

The shift toward moral vices was substantial in the primary database. Statistical models revealed a 52 percent increase in lyrics related to degradation. Themes of harm increased by 49 percent, while cheating rose by 48 percent. Subversion also saw a notable increase of over 40 percent.

In contrast to these rising vices, the moral virtue of care experienced a 24 percent decline over the analyzed period. The smaller Billboard dataset confirmed these general patterns. The popular hits showed an even steeper 70 percent increase in expressions of cheating.

The authors found that these moral expressions were strongly tied to specific emotions and themes. Songs scoring high in care and loyalty tended to feature positive sentiments, expressing joy and themes of love. Lyrics scoring high in harm, cheating, or subversion correlated strongly with negative emotions, including sadness, disgust, and anger.

The researchers also took a brief look at specific historical periods known for social unrest. They compared music from the Vietnam War era to music released during the War on Terror. They found that expressions of degradation, cheating, subversion, and harm were significantly higher during the War on Terror period compared to the Vietnam War era.

When the researchers tested their computer models on specific music genres, they found that certain moral themes were easier to predict within particular styles of music. The concept of purity was most strongly detected in religious music. Themes of harm and degradation were highly predictable in metal music, which often embraces controversial and rebellious topics.

“Over the past six decades, the moral themes in popular lyrics have shifted: themes of harm, cheating and rebellion appear more often, while expressions of empathy and decency have declined, alongside a broader rise in negative emotion,” Saitis said. “We found differences in the moral expression of song lyrics linked to musical style and storytelling traditions, but also to (attributed) artist gender, though these should be interpreted in light of the binary artist gender classification and the substantive artist gender imbalance in the datasets we used.”

As Saitis pointed out, the study categorized artist gender using a strict male or female binary based on available database information. Female artists tended to express slightly higher levels of care in their lyrics, while male artists were more likely to perform songs featuring themes of harm and degradation. However, the binary classification system fails to capture non-binary identities, which limits the interpretation of these gender-based patterns.

The authors note some other limitations to their approach. The analysis relied heavily on Western pop music recorded in English. This focus restricts how these findings might apply to other cultures with different musical and moral frameworks.

The primary dataset also experienced a sharp drop in available songs after the year 2010. This means the trends observed in more recent years might be influenced by a lack of data rather than genuine cultural shifts. The Billboard dataset introduces its own bias, as it reflects only commercially successful tracks shaped by industry marketing and radio play.

Additionally, the machine learning models used to read the lyrics are not perfect. Complex human expressions involving irony, slang, and metaphor can sometimes confuse computer algorithms. The authors note that their research highlights mathematical correlations but cannot prove cause and effect.

It remains unknown whether music actively shapes the moral values of society or simply reflects the changing attitudes of the listening public. “More generally, the key thing to keep in mind is that this work describes what songs express, not what listeners believe,” Saitis said. “Enjoying a song does not mean sharing its moral outlook, but it could influence it, something we intend to study next, including extending the analysis to musical audio signal data.”

To allow the public to explore this technology, the researchers launched a web app based on the AI language model used in their study. Users can upload CSV or Excel files containing text or lyrics, and the tool will automatically assign probability scores across the 10 moral categories.

The study, “Evolution of moral expression in song lyrics,” was authored by Vjosa Preniqi, Andreas Kaltenbrunner, Kyriaki Kalimeri, and Charalampos Saitis.