Popular Song Lyrics Use Fewer Positive and Social Words Over Time

The Sex Pistols made one studio album, Willie Nelson 78; about 30 trained research assistants checked every artist’s discography against the lot. When psychologists then tallied the words in 38,467 songs by 230 popular artists, the share given over to emotional and social vocabulary mostly drifted down rather than up, both as individual careers rolled on and from one generation of debuts to the next. The expectation had been the opposite. Whether that says anything about the artists, or only about the business of selling songs, is harder to pin down.

Logan Gibson, a research associate at Michigan State University, and his colleagues drew their artists from the top 100 of any of six all-time rankings, then gathered lyrics from 3,531 studio albums released from 1947 to 2024. Each song went through a dictionary-based word counter, which reports the share of its words that fall into categories such as positive emotion, anger, sadness, family and friends.

Separating a Career From an Era

Plenty of earlier research has asked whether music changes across decades, while far fewer studies have followed the same artists through their own careers. This one did both at once, fitting multilevel growth curve models that estimate change within each artist alongside differences between artists who debuted in different years, with the debut year of an artist’s first album standing in for the era they came up in. “By looking at both at once, we can begin separating changes that happen within artists from broader shifts in culture and the music industry,” said Jan Brascamp, co-author and associate professor in the Department of Psychology. The separation is statistical, mind you, and the design is observational, nobody assigned artists to careers, so every pattern here is an association.

So are lyrics just getting gloomier? Hardly, or not obviously: negative emotion and sadness words fell in both comparisons, and positive emotion words fell across generations.

Within careers, family words fell furthest, with friends and social references overall dropping too, while anxiety words ran the other way and crept up. Across debut years, newer artists used fewer positive emotion, family and friend words and more anger words, while anxiety and overall social references showed no clear generational trend. The within-career dip in positive emotion, for what it’s worth, was the shakiest finding of the lot (its uncertainty range reached zero), so that one deserves a squint.

Gibson set out the expectation the data failed to meet: “People often become more emotionally regulated and invest more in meaningful relationships as they age, so we might expect that lyrics become more positive or relational across an artists’ career,” he said. He added: “Instead, positive emotions and social references declined.”

Genre Does Not Obviously Explain the Drift

Genre was the obvious suspect, since rock supplied 32.76 percent of the songs and pop 21.23 percent, with country at 12.22 percent and hip-hop and rap at 6.12 percent trailing behind. So the team ran 192 possible comparisons, eight word categories across 12 genres and two kinds of change, to see whether any trend depended on genre. Just 13 came out significant, a hit rate of 6.77 percent against the 5 percent that chance alone would be expected to hand out, so genre does not obviously explain the drift. A word counter is also a blunt instrument: it tallies the share of words in each category, so a song heavy on love words scores the same whether the love is sincere, sarcastic or metaphorical, and lyrics are often more symbolic than literal. Nor are lyrics a diary, since many performers do not write all their own material, which is why the authors treat them as a trace of culture rather than a window onto an artist’s inner life.

Is the story really as neat as the press release makes it sound? It says differences between eras were generally larger than changes within careers; the paper’s own figure bears that out for positive emotion, negative emotion, sadness and anger, but family words, overall social references and anxiety shifted more within careers, and friends came out roughly level.

Gibson reads the declines as a hint that lyrics may be shaped heavily by artistic, cultural or industry forces; that is his inference, and the paper itself says the data cannot determine why lyrical content changes, with psychological development among several equally plausible explanations. The sample doesn’t help either: about 74 percent of songs came from artists in the United States and 18.4 percent from the United Kingdom, and most artists were men, so nobody should assume the pattern travels to other musical cultures.

A Record of What Artists Choose to Sing About

William Chopik, a co-author and former associate professor in the Department of Psychology at Michigan State, sees songs as more than entertainment. People use them, he says, to regulate their emotions, work out who they are and feel connected to others. That makes lyrics a record of what artists, and perhaps their audiences, choose to talk about. So do large-scale declines in emotional and social language point to something bigger: has the way we express feelings and relationships changed more broadly?

A direction, not a reason. Why the feeling words are thinning, or whether the feeling itself is, remains open.

FAQ