Spotify as Archive and Site of Media Archaeology
Emily Van Ryn
How do you remember your life? Through diaries? Through photos? If you’re like most people, you might find your memories tied to music. Whether it’s an album that you can’t separate from the memory of your high school girlfriend, or a song that reminds you of that one day at the lake, recorded music is a powerful tool for memory: an archive. Traditionally, archives are places that store collections of primary source historical records that have been deemed worthy of preservation by an institution. These collections can include practically any form of media including books, film reels, and even entire paintings. On a more personal level, our ways of listening to and storing music have changed drastically over time. Where we once collected records, hoarded homemade mixtapes, or stockpiled CDs, we now have the ability to store countless hours of music in a single smartphone or computer application.
One of the most recent developments in music storage and listening formats is the creation of music streaming services such as Pandora, Tidal, and Apple Music. Spotify is currently the most popular audio streaming service, boasting 345 million users across 170 countries (“Company Info”’). Spotify has been an archive of my own life experiences since my sophomore year of high school in 2015. Though I have an impressive 155 public playlists alone, the entire platform is home to over 4 billion playlists as of 2020 (Iqbal). Spotify allows users to create their own playlists but it also curates playlists of its own, some tailored to individual listeners based on their behavioral data and others curated for the entire platform to suit a particular mood or to provide an overview of a genre of music. In these practices, Spotify is what Jussi Parikka would call a site of media archaeology. Where typical archaeology involves excavating physical sites, media archaeology is interested in examining media of the past to understand the present and the future. In his 2012 book What is Media Archaeology?, Parikka writes that media archaeology is “a set of theories, methods and ways to understand the mediatization of cultures of memory as well as the dynamics of old and new media.” Media archaeology draws on the work of cultural and critical media theorists such as Michel Foucault and Walter Benjamin, as well as entire fields of study – film studies being just one example. Media archaeology argues that we use new media to mourn and reshape old media and that media does not die so much as it transforms. Parikka writes that “some of the reuses and communities – for example, around cassettes – have found a new life with the internet and on smartphones and [iPads].” Here, Parikka is discussing the ways in which the culture and rituals around cassette tapes have transformed and adapted to exist as a digital media. Spotify exemplifies this transformative shift from physical mixtapes into digital playlists.
This paper acts as a case study of how archives are changing in the digital era, where I argue that Spotify acts as both a digital archive and a site of media archaeology while revealing the inherent value of playlists, regardless of their material form. I have identified three primary playlist types on Spotify: Data Driven, User Created, and General. I will begin by examining the work of Jussi Parikka to explore the concepts of media archaeology in relation to data driven and general playlists. I will then review Walter Benjamin’s concept of aura to consider what might be lost in the transformation of homemade physical mixtapes into user created digital playlists, before exploring why playlists are still important to us nonetheless.
Spotify as a Site of Media Archaeology: Data Driven and General Playlists
Spotify is known for its wide variety of playlists and has received considerable notoriety for its data-driven Discover Weekly and Daily Mix playlists, which are built specifically for individual users. In her 2017 Medium piece “How Does Spotify Know You So Well?,” software engineer Sophia Ciocca dives into Spotify’s use and combination of three different song recommendation models and how effective they are in creating these beloved Discover Weekly playlists. These song recommendation models were an efficient move away from older models that relied on the manual filtering and tagging of a song’s attributes. As Ciocca explains, the Discover Weekly playlist employs collaborative filtering, natural language processing (NLP), and audio models. Collaborative filtering looks at implicit feedback, which includes a song’s stream counts, if a user has saved the song to a playlist, and if a user went to the artist’s page after listening to the song. It then compares this information across users, recommending songs to individuals based on their similarity to other users. NLP gathers data from blog posts about songs and artists to more or less tag a song’s attributes in a more efficient and scientific way. Audio models examine what is essentially the fundamental sound of a song – features including time signature, tempo, and key (Ciocca).
Discover Weekly playlists were released in 2015, and the positive response from users drove Spotify to “rethink its focus,… invest more resources into algorithm-based playlists,” and release Daily Mix playlists just a year later (Ciocca). A user can receive up to six Daily Mix playlists a day with the intent of providing personalized playlists that span the broad range of genres a given user might like. For example, rather than shoving Broadway songs into a playlist with hyperpop, there would be two different playlists that cater to these distinctive genre interests while still providing something cohesive. These playlists show that Spotify is a site of media archaeology already: examining past music consumption to understand present and future consumption.
Spotify has innumerable general playlists, those that were not built for individual users based on their behavioural data. Among these are brand-curated playlists, but the commercial aspect of Spotify playlists is beyond the scope of this paper. General playlists can be found under the ‘Browse’ feature of the application where there are over 50 major playlist categories including K-Pop, Sleep, and Latin. Within each category there are hundreds of playlists: the Latin category in particular houses around 500 different playlists (Iqbal). Another category of general playlists is Decades, which include playlists with names like Rock Classics, All Out 50s, Classic Oldies, and so on.
The Decades category is somewhat unique for its subcategories which each contain about 20 playlists. They include playlists consisting of the top hits of each year in that decade, along with other playlists that are genre or mood-specific (80s Rock Anthems and Walking on Sunshine are two playlists found under the 80s subcategory). Playlists within the Decades category often have a description that provides some music historical context. Featuring Lady Gaga on the playlist cover art, the description for All Out 00s says the playlist is made up of “essential tracks from the decade that catapulted electronic music, hip hop, and indie rock into the mainstream” (Spotify). Spotify does not claim to be a music historical or archival service, but its creation of playlists such as these, each one having hundreds of thousands if not millions of followers and account for a third of all user listening time, prove the platform to be a significant player in shaping, preserving, and disseminating music history (Iqbal).
I do not believe Spotify has any ill intent in its creation of decade playlists nor do I believe they have any agenda of erasure. That said, it is worth keeping in mind John Guillory’s words in “Genesis of the Media Concept” where he notes that the English philosopher Francis Bacon celebrated print media because it was “destined to ‘unmask and dethrone’ the tyranny of priests and kings” (325). These priests and kings had control over the dissemination of information and thus shared memory and shared history. Though Spotify likely is not out to rewrite or shape all of music history to their benefit the way priests and kings once did, it is important in this time of rapid change to be mindful of sites of media archaeology like Spotify, as the findings and results of their archival work can shape the future and, in some ways, write our histories.
Spotify as Archive: User Created Playlists and the Question of the Aura
There are a few interesting playlist settings that contribute to Spotify’s overall archival and media archaeological functions. Public playlists can be followed, saved, and sent to anyone. Private playlists can only be viewed by the creator unless they are set to collaborative, where they are then able to be saved by those with access to the playlist link (“Shared Private Playlist”). The collaborative playlist setting also allows for anyone with the playlist link to add songs to the playlist — a common practice for creating party playlists. Collaborative playlists are unable to be public, a measure taken to prevent people from hijacking and ‘vandalizing’ user playlists (“Collaborative Playlist”). As for their archival properties, user created playlists remain on the user’s account unless the user deletes them or they are reported and taken down (“Report a Playlist, Image, or Text”). For me and other Spotify users, this means the playlists of our lives are just a scroll and click away. Playlists for every crush, every friend’s birthday party, and every trip are archived in my playlists tab, where they could very well remain until the end of my time with a Spotify account.
As previously stated, my playlists go back to my sophomore year of high school, meaning that many of them have titles that are embarrassingly angsty (“let me be emo for 1 sec ok”) or funny to rediscover upon realizing which crush they were made for (“april showers” for someone whose birthday was in April). Interestingly, I have a playlist from 2016 called “heard it on the grape vine” which is a collection of songs I first heard on the now defunct video sharing app Vine and a playlist I made in 2020 called “tiktok is my latest disease” which catalogues songs that have gone viral on TikTok, another video sharing app that is considered to be an evolution of Vine. Most of my older playlists are now organized into playlist folders, tucked away to minimize clutter in my playlist tab. Spotify accounts are digital archives, the user created playlists are what is archived, and within them are the memories attached to each song.
What are you really giving someone when you make them a playlist?
It is important to acknowledge the reality that Spotify users do not own the playlists they create the way they would own a physical mixtape. Spotify may not own the music it licenses, but it does own user behavioural data and user created playlist configurations (Eriksson et al.). This calls into consideration what is lost between a physical mixtape, which could potentially be kept forever, and a digital playlist created on a music streaming service that could be deleted by the platform’s owner at practically any time. What are you really giving someone when you make them a playlist? Nothing truly tangible. You send them a link, but ultimately nothing about this gift has any basis in the material world.
Walter Benjamin offers insight into the significance of this immateriality in his 1935 essay The Work of Art in the Age of Mechanical Reproduction, writing that upon not being able to share physical space with a work, “the quality of [a work of art’s] presence is always depreciated” speaking to the ways in which lacking physical access to art or art objects such as mixtapes creates a certain loss of meaning (4). This meaning is what Benjamin calls the aura of a work, the particular feeling and experience of beholding an original art object in person, what he defines as a “unique phenomenon of a distance however close it may be” (5). The aura has a “unique value” that has “its basis in ritual, [in] the location of its original use,” again emphasizing the importance of the tangibility of a work, even if convention or policy prevents you from literally touching it (6). The same way there is something incredibly special about being close to an original painting, there is also something special about holding a physical mixtape made for you by someone you care about.
If this special aura is lost in the shift from a physical mixtape to a digital, immaterial playlist, what value do playlists hold? Why are there 4 billion of them on Spotify and why do I have over 155 of them? Why are playlists still important to us? It may not be commonplace to hand-make a mixtape anymore but the gesture of curating and customizing a digital playlist is not only meaningful in its own right but is ultimately a more accessible way to enjoy and share music. Benjamin writes that “mechanical reproduction,” in this case the creation of a digital rather than physical playlist, “emancipates the work of art from its parasitical dependence on ritual” (6). The immateriality of digital playlists makes it much easier to create, share, customize, and even update playlists. This lack of physical constraint allows digital playlists to enable connection at speeds and distances that traditional physical mixtapes just can’t. Mixtapes allowed creators to engage with the physicality of these musical archival objects by adding decorations, handwritten tracklists, and other details that contributed to the overall aura of the object, but ultimately their transferability and collaborative capabilities are relatively limited. Spotify gives users the ability to customize the title, description, and cover art of their playlists, allowing them to rearrange and continue to add songs to a playlist long after the link has been shared. These playlists can be made collaboratively, allowing for meaningful connection from a distance. Coming up with clever titles based on inside jokes, designing playlist cover art from a funny photo of your friend, and arranging songs in a playlist to create a particular flow are all thoughtful acts that make for a special, albeit digital, art object.
Conclusion
Spotify is a platform that functions as both an archive and as a site of media archaeology. This intriguing combination speaks to the rapid change to what ‘archive’ means and what qualifies as one in the digital era. As major music streaming platforms like Spotify continue to delve deeper into data aggregation and analytics to shape what they show users, we must be mindful of the role these platforms play in the shaping and recording of histories, even if it is in the form of seemingly trivial playlists. Music is a large part of how we remember our lives, our relationships, and the past in general. Despite the transformation of physical mixtapes into digital playlists, this archival work remains a meaningful practice. We should be sure that the archival and archaeological nature of platforms like Spotify stay in a place where they serve us and help us to preserve our ever-changing media practices.
References
“All Out 00s.” Spotify, Spotify, Accessed 5 Mar. 2021, https://open.spotify.com/playlist/37i9dQZF1DX4o1oenSJRJd?si=Ni9UTGHnRo6zhwwke2dhw
Benjamin, Walter. The Work of Art in the Age of Mechanical Reproduction. Translated by Harry Zohn, 1935.
Ciocca, Sophia. “How Does Spotify Know You So Well?” Medium, Medium, 5 Apr. 2018, https://medium.com/s/story/spotifys-discover-weekly-how-machine-learning-finds-your-new-music-19a41ab76efe.
Eriksson, Maria, Fleischer, Rasmus, Johansson, Anna, Snickars, Pelle, and Vonderau, Patrick. “Where Is Spotify?”, Spotify Teardown: Inside the Black Box of Streaming Music, MIT Press, 2019.
“Company Info.” For the Record, Spotify , 31 Dec. 2020, https://newsroom.spotify.com/company-info/.
Emily Van Ryn is a writer and artist originally from Minneapolis, Minnesota. Emily’s interests lay in both the spiritual and emotional experience of non-normative identity in hegemonic societies.