How Sad Are the Sad Girls?
Challenging the Indie Music Trope with Data
“Sadness made one 'interesting.' It was a mark of refinement, of sensibility, to be sad. That is, to be powerless.”
- Susan Sontag, Illness as Metaphor, 1978
Two years ago, Jenn White of NPR’s 1A podcast declared 2021 “The Year of the Sad Girl” (White, 2021). She was referring to the influx of highly emotional and poetic music that had recently been released by female musicians. Jia Tolentino of The New Yorker also noticed this in 2018 when she wrote one of the first profiles on these quote-on-quote “sad girls.” Her article, titled, “The Wry Young Women Writing Sad, Buoyant, Beautiful Songs,” lauded artists like Soccer Mommy, Lucy Dacus, and Snail Mail’s Lindsey Jordan for their self-consciousness, lyrical vulnerability, and emotional accuracy (Tolentino, 2018). It was around this time that I had also started listening to their music. Like Tolentino, I identified with many of the themes in these artists’ lyrics. The songs might have been heavy, but I found solace in hearing someone like me sing about their romantic relationships, queerness, and mental health.
Five years later, the phrase “sad girl music” has come to refer to a subgenre of indie music to which artists like Phoebe Bridgers, Mitski, Julien Baker, Jay Som, and Japanese Breakfast are often categorized. At its most empowering, the phrase functions as a calling card for young girls online to announce their music tastes and develop a community of like-minded friends. For the artists, this has sometimes resulted in hyper-devoted fanbases. Phoebe Bridgers fans, for instance, call themselves “Pharbz” and have developed a reputation for fainting from exhaustion during her performances after camping out at the venues beforehand. Similarly, as the popularity of these artists grows, the music industry and mainstream pop artists have picked up on the marketability of sadness. When Taylor Swift re-released her album “Red” in 2021 she also released a muted piano version of her infamous break-up song called “All Too Well (Sad Girl Autumn Version),” not-so-cheekily referencing the TikTok phrase trending at the time. When used most insidiously, the term “sad girl music” reduces the range of these artists’ songwriting and musical talents to an all-too familiar stereotype of the sad, emotional woman.
Often regarded as the figureheads of the “sad girl music” scene, Lucy Dacus, Phoebe Bridgers, and Julien Baker (who together comprise the supergroup Boygenius) have been vocal about their dislike of the term. In an interview with The Pitchfork Review, the trio points out that discussions of “sad girl music” often equate overall emotionality with sadness, which disregards the complexity of someone's lived experience (Patel, 2023). Take for example, nostalgia, which is a common theme sung about by these “sad girl” musicians. Nostalgia may naturally be tinged with sadness, but this does not mean it is an intrinsically sad emotion. Rather, the potency of nostalgia’s specific type of sadness comes from the realization that what you long for is out of reach. Likewise, forgiveness embodies this duality. To forgive implies past pain, but it also indicates moving beyond it. Lucy Dacus has said that this sentiment of getting better is a driving force in her songwriting (White, 2021). This duality is clearly at play in “Night Shift,” a searing and soaring breakup anthem in which Dacus settles into the fresh pain of a breakup but also finds comfort in knowing that one day the pain will fade:
You got a 9 to 5, so I'll take the night shift
And I'll never see you again if I can help it
In five years I hope the songs feel like covers
Dedicated to new lovers (Dacus, 2018).
Now, Dacus describes “Night Shift” as a victory song, saying it is not about being hurt in the past, but rather about the present (White, 2021). I can attest to this. At her shows, the crowd screams her lyrics along with her. We know that we have all been hurt, but we also recognize that we have all healed. Her lyrics may be sad, but Dacus’s performance is radically cathartic.
In each of these examples, sadness is just one facet of these artists’ songwriting. Dacus believes that portraying women as sad makes them less of a threat. As a result, a woman’s happiness becomes something to be fulfilled by someone else, which ultimately crafts a space of usefulness for men (White, 2021, 27:32). This is the same sentiment reflected in this essay’s opening quote from Susan Sontag (“It was a mark of refinement, of sensibility, to be sad. That is, to be powerless.”) (Sontag, 1978). In her interview with Jenn White, Dacus discusses a tweet she posted which states:
sadness can be meaningful but I got a bone to pick with the “sad girl indie” genre, not the music that gets labeled as that, but the classification and commodification and perpetual expectation of women’s pain, also I don’t think my songs are sad, anyways good morning ☀️(Dacus, 2021)
Dacus says the tweet is inspired by Leslie Jamison’s 2014 essay titled “Grand Unified Theory of Female Pain,” in which Jamison examines her personal relationship to pain and relates it to how art and media has reacted to women’s suffering (Jamison, 2014).
Jamison’s essay reflects themes that Julien Baker discusses with Laia Garcia-Furtado in an interview for Harper’s Bazaar. Baker believes sadness arises when women and queer people realize society has excluded them from certain opportunities and spaces. In 2021, the CDC published their Youth Risk Behavior Survey, which reported on data collected on teenagers between 2011 and 2021. The study found that “Female students and LGBQ+ students are experiencing alarming rates of violence, poor mental health, and suicidal thoughts and behaviors” (Centers for Disease Control and Prevention, 2021). Teenage girls have attributed this to “growing up in a social media culture, with impossible beauty standards, online hate, academic pressure, economic difficulties, self doubt and sexual violence” (St. George et al., 2023). Specifically, of all high school students, 42% experienced persistent feelings of sadness or hopelessness. When looking at the data for female and queer students, the rates are even higher. 57% of female students and 69% of LGBTQ+ students experienced these feelings, compared to 28% of male students and 35% of heterosexual students (Centers for Disease Control and Prevention, 2021, p. 60). In contemplating the recent virality of “sad girl music,” the CDC data proves that this sadness is a very real facet of the lived experience of today’s female and queer teenagers. When discussing her thoughts on the genre, Dacus has similarly expressed that “whatever is resonating with people is indicative of how people feel in the moment” (White, 2021, 30:47).
One of the key themes in Jamison’s essay is that while sometimes the sadness conveyed through writing, art, and music can feel cliche, trite, or exaggerated, it does not mean the experience is any less real for the author. She concludes her essay stating:
The wounded woman gets called a stereotype, and sometimes she is. But sometimes she's just true. I think the possibility of fetishizing pain is no reason to stop representing it. Pain that gets performed is still pain. Pain turned trite is still pain. I think the charges of cliché and performance offer our closed hearts too many alibis, and I want our hearts to be open. I just wrote that. I want our hearts to be open. I mean it. (Jamison, 2014, p. 128)
While the phrase “sad girl” may trivialize the vast emotional landscape experienced by these artists, there is clearly a reason as to why teenagers and young adult listeners have identified with their music.
In many ways, the “sad indie girls” of today are situated within a tradition of personal and political songwriting established by female songwriters of the late 20th century. The Riot Grrrl movement of the early 90s sought to dismantle patriarchal structures through radical feminist punk music (“Riot grrrl MANIFESTO”, 2013). Sarah McLachlan formed Lilith Fair as a challenge to the music industry when she was told by executives that an all-female festival lineup would never sell (Weiner, 2022). Like today’s “sad girls,” the artists of Lilith Fair and the Riot Grrrls were also often met with derision from media publications. Critics disparaged Lilith Fair as “mom music” (Weiner, 2022), and the Riot Grrrls were “made out to look like we were just ridiculous girls parading around in our underwear” (Huber, 2010, p. 69).
Ultimately, the derisive terms used to describe female musicians, including yesteryear’s feminist songwriters and today’s “sad indie girls,” comes from a long, misogynistic history of undervaluing womxn’s experiences. Any criticism of these artists, regardless of reductionary labels like “sad girl music,” instead calls attention to the urgent need to change the way we speak about the emotional difficulties women, queer, and non-binary folks experience.
In her essay, Jamison says that the pain portrayed in lyrics by 90’s artists such Tori Amos, Fiona Apple, and Mazzy Star acted like “scars to try on like costumes” (Jamison, 2014, p. 121). These songs provided Jamison with an opportunity to empathize with these writers, even if she did not directly experience their pain. Likewise, Julien Baker sees the positive side to the struggle of deprivation and exclusion experienced by young girls and queer teenagers. She states, “...if you allow yourself to be unguarded about that and feel the natural compassionate response—it is pain, right? Like, it’s regret and sadness. And that’s an okay thing to feel. Because if the pain is instructive, the sadness is instructive about how to make a more compassionate world” (Garcia-Furtado, 2022). Baker does not deny the sadness, she embraces it. Because if we have been sad and we reflect on it and we sing along with it in a crowded concert venue filled with others who have also been sad, I too believe that we can ultimately advocate for this change.
Data & Methodology
The primary goal of this project was to compile a dataset that presents the multi-dimensionality of the contemporary “sad girl indie music” genre. I believe that the emotionality these artists are portraying is far more nuanced than the term “sad girl music” implies. I wanted to develop a tool that facilitates the discovery of these artists and provides a starting point for further analysis of their music. Through quantitative and qualitative data, I want to uncover common themes and song attributes that unite these artists with the goal of ultimately debunking the stereotypes surrounding their music. My resulting dataset was compiled using Python scripts that collected data from the Spotify API and Genius API. Additional text analysis was provided by the OpenAI API.
Creating the Sample Set
To begin, I developed a preliminary list of artists who I believe are representative of the “sad girl indie” genre. This list was based on a combination of internet research and my own personal listening habits. The first section of my list features artists who are most commonly associated with the genre in music publications and online discussions. Each of these artists are predominantly influenced by rock and folk music and include:
Lucy Dacus
Phoebe Bridgers
Julien Baker
Mitski
Snail Mail
Soccer Mommy
Waxahatchee
Japanese Breakfast
Adrienne Lenker
Julia Jacklin
Angel Olsen
Sharon van Etten
Jay Som
Recently, I have also noticed a second generation of artists who are working with the same emotional ranges as these aforementioned “first-generation sad girls." However, these artists are influenced by a wider variety of genres including bedroom pop, alternative R&B, and experimental music. These artists are typically in their late teens or early 20s, and have developed their audiences through social media platforms like TikTok, Bandcamp, and SoundCloud. This portion of the list includes:
Clairo
mxmtoon
girl in red
beabadoobee
Samia
Wallice
BLÜ EYES
Orla Gartland
Renée Rapp
Indigo De Souza
Arlo Parks
Remi Wolf
WILLOW
Finally, I wanted to include additional artists who I personally enjoy. As a queer, young woman, I think my own experience relating with artists in this subgenre can inform the analysis. After all, music is both highly personal, yet also communal. The final portion of my list includes:
Maggie Rogers
Kate Bollinger
Becca Mancari
Wilma Laverne Miner
Margaret Glaspy
Jade Bird
Ada Lea
Delaney Bailey
Angie McMahon
Natalie Prass
Dora Jar
Frankie Cosmos
Cornelia Murr
Grace Ives
This preliminary list resulted in 41 total artists. Once it was created, I used the Spotify API to retrieve each of these artist’s Related Artists. This resulted in a list of 431 unique artists.
At this point, I began narrowing down this list of artists to those who are female-identifying or non-binary. I determined an artist’s gender identity through manual research online. It is important to acknowledge that a certain level of assumption had to be made when I was selecting artists. When I could not find an artists’ self-volunteered pronouns (typically through verified Instagram profiles or self-published Genius biographies), I deferred to how news articles and Wikipedia pages used an artist’s pronouns. Since gender identity is fluid, an artist’s pronoun usage may have evolved since I created this sample set.
When narrowing down the sample artists, I also chose to ignore mainstream musicians such as Taylor Swift, Lorde, Billie Eilish, Olivia Rodrigo, and Lana Del Rey who are commonly associated with the “sad girls.” While I believe these artists’ music is meaningful to consider in the context of this study, I was worried about the frequent commercialization and marketing of the trope in these artists’ public images. Additionally, when I included these artists in the preliminary sample, my data skewed towards general pop, and I wanted this analysis to focus primarily on independent and up-and-coming artists. Finally, I also chose to exclude collaborative projects and bands where it was not evident the entire musical group was female-identifying or non-binary. This may have excluded some interesting work being created by leading front-womxn. Ultimately, the final sample set includes 237 artists who are aligned with the “sad girl indie” music genre.
Data Collection & Visualization
Once my sample set had been completed, I began pulling data from the Spotify and Genius APIs. I wanted to collect both quantitative data about the artists’ songs and qualitative data about the lyrics. The questions I hoped to answer through this data included: How does the way a song sounds inform its sentiment? Do sadder artists tend to write songs in one key over another? What is the overall energy of this music? What are the common themes and emotions these artists are writing about?
To begin, I first retrieved general data about each of the artists. This included biographies from Genius and data from Spotify such as the genres they are classified, their popularity, and top ten trending tracks. Next, to collect quantitative data about the artists’ tracks I made two bulk data calls to the Spotify API using their Get Track and Get Track Features endpoints. For each song, this data includes a popularity score between 0 and 100; whether a track was explicit; its duration; and attribute data including danceability (how suitable a track is for dancing), energy (the overall intensity of a song), key, instrumentalness, liveness, loudness, mode (ie. major or minor key), speechiness, tempo, time signature, and valence (how happy or sad a song sounds). I was specifically interested in the danceability, energy, and valence metrics which essentially measure the emotional feeling of a song.
Next, to collect qualitative data I scraped the lyrics for each artists’ top ten songs using the LyricsGenius Python client. Through this lyric data, I hoped to discover common themes that exist amongst these artists. To do so, I used OpenAI’s chat completion tool to summarize each of the songs in under 50 words. I then passed these summaries a second time into the chat completion tool, asking specifically for keywords from the first result. After cleaning the resulting list using OpenRefine, I was left with 2,367 unique keywords found in my sample set’s lyric data. The top keywords sung by these artists include “love,” “relationship,” “struggling,” “frustration,” “desire,” “moving on,” “uncertainties, “longing,” “regrets,” and “hope.”
Finally, the retrieved data was compiled into CSV files which were imported into Tableau for visualization. The resulting product is a dashboard that allows users to explore the qualities of each of the 237 artists’ discographies. Metrics are displayed for each artist’s lyrical keywords, track attribute averages, top tracks, popularity, explicitness, and song keys. For a walk-through of the Python code used to compile and prepare this data, please refer to the project’s ReadMe, available at: https://github.com/alysedelaney/how-sad-are-the-sad-girls
Conclusion: Next Steps
In the future, I hope this dataset can be used to perform an in-depth analysis of these artists’ music and the “sad girl” genre as a whole. Through an initial exploratory analysis of the data, I can see support for my hypothesis that the emotional ranges these artists are capturing are far more diverse than the term “sad girl music” implies. For instance, Phoebe Bridgers, an artist well-known for her melancholic tone, may sing about themes such as “addiction,” “emotional pain,” “grief,” and “sadness,” but she also sings about “adventure,” “dreaming,” “glimmer,” “nostalgia,” and “moving on.”
This data has the potential to illustrate how genre classification and sentiment analysis tools may be operating on patriarchal standards that inherently skew female and non-binary artists’ metrics towards the negative end of the spectrum. Regression analysis could reveal insights between the relationships between track attributes and how Spotify assigns these metrics to songs. For instance, if a song has a slow tempo, does that by default produce a low valence score? Additionally, I would be interested in exploring how listeners experience music from a phenomenological perspective. For instance, how might a listener’s personal memory of a song inform their sentimental perception of it? Can a song measured as “sad” still evoke happy feelings? I believe this research could be used to reveal shortcomings in sentiment analysis tools that fail to take into account the variation and multiplicity of experienced emotions.
References
Centers for Disease Control and Prevention. (2021). Youth Risk Behavior Survey, Data Summary & Trends Report, 2011-2021. Division of Adolescent and School Health. https://www.cdc.gov/healthyyouth/data/yrbs/pdf/YRBS_Data-Summary-Trends_Report2023_508.pdf
Dacus, L. [@lucydacus].(2021, February 27). sadness can be meaningful but I got a bone to pick with the “sad girl indie” genre, not the music. [Tweet]. Twitter. https://twitter.com/lucydacus/status/1365662606925651968
Dacus, L. (2018). Night Shift [Song]. On Historian [Album]. Matador Records.
Garcia-Furtado, L. (2022, February 9). The Joy of Sad Girl Music. Harper’s BAZAAR. https://www.harpersbazaar.com/culture/features/a38647867/the-joy-of-sad-girl-music-february-2022/
Huber, J. L. (2010). Singing it out: Riot grrrls, Lilith Fair, and feminism. Kaleidoscope: A Graduate Journal of Qualitative Communication Research, 9(5).
Jamison, L. (2014). Grand Unified Theory of Female Pain. The Virginia Quarterly Review, 90(2), 114–128. https://www.jstor.org/stable/44714543
Patel, P. (Host). (2023, March 30). Boygenius Interview (S2 E22) [Audio podcast episode]. In The Pitchfork Review. Pitchfork. https://podcasts.apple.com/us/podcast/boygenius-interview/id1516844211?i=1000606568341
Riot grrrl MANIFESTO. (2013, April 27). Actipedia. https://actipedia.org/project/riot-grrrl-manifesto
Sontag, S. (1978). Illness as metaphor. Farrar, Straus and Giroux.Tolentino, J. (2018, April 7). The Wry Young Women Writing Sad, Buoyant, Beautiful Songs. The New Yorker. https://www.newyorker.com/culture/culture-desk/the-wry-young-women-writing-sad-buoyant-beautiful-songs
Spotify. (n.d.). Web API Reference | Spotify for Developers. Retrieved May 1, 2023, from https://developer.spotify.com/documentation/web-api/reference/get-an-artists-related-artists
St. George, D., Reynolds Lewis, K., & Bever, L. (2023, February 17). The crisis in American girlhood. Washington Post. https://www.washingtonpost.com/education/2023/02/17/teen-girls-mental-health-crisis/
Weiner, L. (2022, July 5). 25 years on, Lilith Fair is a reminder of how one woman’s radical idea changed music. NPR. https://www.npr.org/2022/07/05/1108635464/25-years-on-lilith-fair-is-a-reminder-of-how-one-womans-radical-idea-changed-mus
White, J. (Host). (2021, November 18). 2021, The Year Of The 'Sad Girl' [Audio podcast episode]. In 1A. NPR. https://www.npr.org/2021/11/18/1057011418/2021-the-year-of-the-sad-girl