Fans recognized another new era for Taylor Swift — and the science backs it

A methodological approach used to analyze Swift’s catalog offers a straightforward and easily interpretable way to explore song lyrics.

By Sarah Steimer

“I stayed up entirely too late on the release day of Life of a Showgirl,” says Jean Clipperton, an assistant senior instructional professor in the MA Program in Computational Social Science. But Clipperton wasn’t just staying awake to listen to the new album — she is a fan — she was running it through a text analysis. “It's different from every other album that she has put out, which is what everyone is saying. But it's just really fun to be like, Yeah, there is an empirical basis for this.”

Clipperton and her coauthor Loizos Bitsikokos, a graduate student at Purdue University, recently published a new paper in Psychology of Aesthetics, Creativity, and the Arts, “You Do Not Get to Tell Me About Sad: Swiftian Saudade in Taylor Swift’s Lyrics.” The research uses a lexicon-based approach with principal component analysis to explore the pop star’s music and her evolution as a songwriter.

Clipperton began analyzing Swift’s lyrics in a course she was teaching, opting to introduce a segment on text analysis via the artist’s work, finding it to be a more relatable content source for students. As Clipperton built the curriculum, she found a dearth of modern pop musicians in computational studies literature — and few studied female artists.

The research team also had the benefit of “Swifties” (Taylor Swift fans) frequently sharing their own theories behind the music, which they could test against. As Clipperton put it, Swifties like data.

“Can we think about her work in similarities and differences?” she asked. “Taylor is known for wanting to have a clear plan or orchestration. I just wanted to know if there was a way that we can take a step back and get a bigger, clearer picture of what's happening, and that's what this (analysis) enabled us to do.”

She and Bitsikokos considered patterns in Swift’s songwriting that her fans often highlight: how track five is often the “emotional realness” of the album; and Swift’s prosperity to juxtapose happiness and sadness in her music. The researchers used the phrase “Swiftian Saudade” to describe Swift’s ability to invoke a mixture of sadness, anticipation, nostalgia, and joy in her lyrics, referencing the Portuguese term for longing, melancholy, or nostalgia. In addition to considering the positive and negative emotions in her songs, the researchers also wanted to consider Swift’s progression as an artist.

The team used a principal component analysis in their text analysis model, creating a 2D plane of Swift’s songs and albums, a method that allowed them to identify key themes and visualize patterns in the data. Words or phrases within her songs were categorized as, for example, fearful, trusting, or joyful. The team then plotted songs and albums along the plane, using axes that showed positive or negative dimensions, and whether the content was more lyric- or emotion-focused.

The research empirically showed much of what Swifties believed: Swift’s work has changed over time, moving from a more lyric-focused place toward one of deep and conflicting emotions. The analysis showed this to be particularly true within track five on her albums, with these showing the highest rate of “Swiftian Saudade” at nearly 50% and a significantly higher level of authenticity than her other tracks.

The study also revealed three ‘meta-eras’ of Swift’s catalog: Her early work was more youthful, hopeful, and a bit more positive. Her mid-career work explored her emotions a bit more, but not quite as linguistic as previous albums. Her most recent era — prior to Showgirl — are much more emotionally charged, leaning into more negative emotions. Of course, Clipperton cautions that this is strictly a text analysis, meaning it can't always capture layers of meaning or nuance. 

The research was published prior to the release of Swift’s latest album, but Clipperton ran the lyrics through the analysis, finding it to be different from every other album Swift has put out. Other departures on the album? “Eldest Daughter,” Showgirl’s fifth track, is far more emotionally charged and less narrative-laden than the fifth tracks on her other albums. And her most joyful, upbeat song doesn’t seem to be the usual track 11.

“She has done this before,” Clipperton says. “She moves into a new space, she explores it a little bit, and then she jumps into a different space. I think this is her M.O. … It could be that this is something that's strategic, where this may be a new space that she continues to explore for the next few albums. I wouldn't be shocked because 13 is a very important number for Taylor Swift fans, and the next album is No. 13. I would imagine that she would do something big, and it wouldn't shock me if it were particularly joyful — given where she's at and what she's doing and what that album would mean.”

While the findings validate her theories as a fan, Clipperton says she’s most excited about the methodological contribution that the work brings to the field.

“I was very intentional in the methodological approach that I use, that it's pretty straightforward and easily interpretable,” she says. “I really wanted to be able to generate something that someone could look at and understand.”