Pop Songs Have Grown More Immoral, Meaner Over Past 60 Years

Lyrical Themes Show Rise In Harm and Rule-Breaking

There’s a reason a grandparent’s record collection feels so different from what’s topping the charts today, and it goes beyond sound. Over the past six decades, the words woven into popular music have drifted away from themes of care, loyalty, and purity, and increasingly toward themes of harm, degradation, and rule-breaking. Now researchers have the numbers to back it up.

A study published in Scientific Reports analyzed the lyrics of hundreds of thousands of English-language songs, drawn from two separate datasets, one covering 1960 to 2010 and another tracking Billboard chart hits through 2023, to measure how the moral character of popular music has changed over time. Expressions of moral “vices,” things like harm, cheating, and rule-breaking, have climbed sharply, while expressions of moral “virtues,” like care, purity, and loyalty, have fallen. Researchers also found that these shifts showed up differently depending on whether a song was credited to a male or female artist, and that they varied widely by musical genre, though those gender-based findings come with caveats about how artists were classified and represented in the data.

Song lyrics have long been treated as a kind of cultural mirror, reflecting the values and attitudes of the era that produced them. If that holds true, then the picture coming back over the past several decades is one of shifting moral priorities, at least in the music that becomes popular.

Scanning Six Decades of Pop Music Lyrics

To conduct the study, researchers drew on two collections of song lyrics covering different time periods. First came a database of roughly 2 million songs, from which they focused on English-language lyrics published between 1960 and 2010. After filtering for artists with more than 10 published songs, that pool came to 377,812 songs. A second collection was pulled from the Billboard Year-End Hot 100 charts, the annual list of the most commercially successful songs, covering 1960 through 2023 and producing a final set of 5,580 songs, roughly 87 per year. Both datasets were analyzed separately, and their date ranges do not fully overlap.

To measure the moral content of all those lyrics, the team used an AI language tool built on the same kind of technology behind today’s chatbots, trained specifically to spot moral themes in text. It scored each song across ten moral dimensions, drawn from a framework in moral psychology that splits values into five pairs, each with a “virtue” side and a “vice” side: Care versus Harm, Fairness versus Cheating, Loyalty versus Betrayal, Authority versus Subversion, and Purity versus Degradation. Each song got a score between 0 and 1 on every dimension, with higher numbers meaning a stronger presence of that theme.

By the Numbers: A Six-Decade Moral Shift

What emerged was a consistent pattern across both datasets. Vice categories, including Harm, Cheating, Subversion, and Degradation, all rose over time. By the study’s models, Degradation changed by roughly 52% in the larger dataset over the study period, Harm climbed about 49%, Cheating rose nearly 48%, and Subversion increased by around 41%. In the Billboard data, the jumps ran even sharper for some categories: Cheating rose about 71%, Degradation roughly 62%, and Subversion around 50%.

Virtue categories largely declined. Care, which captures empathy and concern for others, fell about 24% in the larger dataset and 30% in the Billboard set. Purity, tied to themes of spirituality and sanctity, dropped as well. In the larger dataset, researchers reported that their models explained a large share of the variation in moral scores over time, with release year and artist gender among the main factors they accounted for.

Gender differences ran throughout, though they call for caution. Because the analysis used a binary male/female split and female artists were heavily underrepresented in both datasets, the patterns are suggestive rather than settled. Within that limit, lyrics by female artists more often scored higher on Care, Loyalty, and Betrayal, pointing toward a focus on relationships and emotional connection. Lyrics by male and mixed-gender artists scored higher on Harm, Subversion, and Degradation, especially from the 1980s onward.

In the Billboard data, Harm scores among female artists rose during the late 2010s. Researchers noted that the timing lined up with a period of broad cultural change, but were explicit that the data cannot establish any direct connection, and that it may just as easily reflect shifts in genre trends or lyrical style.

Genre Shapes the Moral Language of Lyrics

Beyond the big-picture trends, the study tested how well an AI model could predict a song’s moral profile from other features of its lyrics, things like emotional tone, topic, and the general feel of the words. Care, Harm, and Subversion ranked among the most predictable, while Fairness proved the hardest to detect, a result that matches earlier research.

When researchers trained separate models for individual genres, the predictions got sharper. A model’s ability to flag Harm and Degradation was strongest when trained on metal lyrics, which fits metal’s long history of dark and rebellious themes. Purity and Authority came through most clearly in religious music. Care was easiest to predict in R&B and Soul/Funk. Genre, it turns out, carries its own moral vocabulary.

Certain lyrical topics also tracked tightly with specific moral dimensions. Songs in the “Love and Emotions” theme lined up with Care and Loyalty. Songs tagged “Violence and Darkness” lined up with Harm, Subversion, and Degradation. Negative sentiment, including anger, disgust, and sadness, ran close to the vice categories throughout.

Pop Music as a Cultural Mirror

Researchers are careful to call the study descriptive, not a verdict. They cannot tell whether darker lyrics are nudging broader cultural attitudes, or whether shifting social values are simply turning up in the music people make and celebrate. That relationship may well run both ways, with music reflecting culture and shaping it at once, though a correlational design cannot prove the point.

Real limits apply. Coverage centers almost entirely on Western, English-language popular music, so the conclusions don’t necessarily travel to music from other parts of the world. That larger dataset thins out sharply after 2010, which means late-period trends there may reflect gaps in the data rather than genuine cultural change, and deserve extra caution. Billboard data captures only commercially successful songs, a filtered slice of what’s popular rather than a full snapshot of everything being made. And the AI tools that scored the lyrics, capable as they are, can miss irony, metaphor, and layered meaning a human reader would catch.

Even so, the scale is hard to wave off. When a pattern this consistent shows up across nearly 400,000 songs spanning more than six decades, and holds in two separate datasets, it’s tough to write off as noise.

Popular music has always been contested ground for culture and identity. What this study adds is a detailed map of how the moral language of popular lyrics has shifted, one song at a time.

---