_{Die deutsche Fassung dieses Texts ist hier zu finden.}

In her excellent book Mood Machine: The Rise of Spotify and the Costs of the Perfect Playlist, Liz Pelly quotes from an essay entitled »On the History of the Playlist,« included in 2020’s Listen to Lists! DNA #2, edited by Lina Brion and Detlef Diederichsen. The published version differs greatly from my first draft, which I have decided to make available separately simply because I like it better.

The following text largely corresponds to this first draft and has been edited, but not updated. Four and a half years after it was written, the last sentence of the text should therefore be taken all the more seriously. What is particularly important at this point is that Pelly has now uncovered the practice behind the so-called fake artists in a chapter of her book, published in Harper’s under the title »The Ghosts in the Machine.«

This is my own translation of the article from the original German. Please note that I have opted to translate all German-language quotes from the original to English even when there are official English-language versions/translations of the cited literature available, as in the case of Shuhei Hosokawa’s The walkman effect.

At the beginning of July 2020, the Wikipedia category »Lists of Music Lists« lists a total of 54 entries, including »Lists of Geordie song-related topics,« »Lists of music inspired by literature« or »Lists of UK Independent Singles and Albums Breakers Chart number ones«.¹ This overview, which is itself structured as an alphabetically sorted list, thus leads to further lists, for example a chronological listing of lists that show the first positions of separate charts for different decades; charts that are themselves generally presented in list form. The category comes with a dry warning: »This list may not reflect recent changes.«²

Lists are, it seems, more complicated than they look. As a rapidly changing form of cultural production, music nevertheless uses them almost fetishistically to create order, construct chronologies and make knowledge quickly consumable. Or it strives to ensure comparability and thus quantitative or qualitative judgements, to construct or even consolidate hierarchies. For these purposes, every list homogenises the heterogeneous, enables a certain form of reception and creates a form of supply that responds to a demand, anticipates it, or even produces one in the first place. The list is therefore always economic(al), because on the one hand it saves time and effort, while on the other it both facilitates and shapes economic interests.

This is why lists are both documents and catalysts of cultural change. Charts, for example, derive their legitimacy from the fact that they depict developments at regular intervals in constant reference to one another. External processes find their expression in the internal alterations of these charts. Cultural events and the social movements associated with them, technological developments, economic fluctuations: All of this can be gleaned from a comparative look at a few lists. However, it is not only the content itself that is subject to change, but also the form of the list itself. The list not only groups together media or media content, it also conveys them and is therefore a medium itself—and as such is just as malleable as what it organises.

The playlist is the most ubiquitous iteration of the list in the 21st century and thus stands at the end of cultural and social, but above all technological and economic developments that it also facilitates. According to Spotify, more and more music was listened to via playlists in the second decade of this millennium, and the company undoubtedly played a large part in this.³ But like the tangled web of lists about music lists, which are listed in a Wikipedia category set up specifically for this purpose, the term playlist is a complex one. After all, what even is a playlist?

Is the playlist »ganz einfach eine Sammlung von Songs [quite simply a collection of songs],«⁴ as Spotify claims on its homepage? A »Wiedergabeliste [playback-list],«⁵ as Wikipedia translates the term into German, i.e. »vorwiegend die Zusammenstellung von digitalen Musikstücken (siehe auch MP3), die auf Computern abgespielt werden [primarily the compilation of digital music tracks (see also MP3) that are played on computers]?«⁶ Possibly, although they are now primarily accessed via the cloud. Does this make the playlist the »album of the streaming world,«⁷ as music journalist John Seabrook claims in The Song Machine. Inside the Hit Factory? Maybe, though probably not, as long as the concept of the album is not tied solely to the unchanging sequence of individual tracks, but also to authorship.

»A playlist is a set of songs selected by human curation or a computer algorithm (e.g., machine learning), or a combination of the two methods,«⁸ writes economist Alan B. Krueger in his book Rockonomics. What the Music Industry Can Teach Us About Economics (and Our Future) and thus defines the playlist not by the act of production, but that of curation. According to K. E. Goldschmitt and Nick Seaver, this is an act that always has a proportion of human and machine components⁹, which is why it is difficult or even impossible to determine a playlist’s authorship. This is especially true because its contents, the individual pieces of music, are generally not produced by the human-machine curators itself. The playlist thus has a diffuse authority but no concrete authorship.

And why are playlists, what is their purpose? Are they a mere marketing tool, functioning as a »word of mouth taken to an industrial scale,«¹⁰ as Chris Anderson characterised them in 2006’s The Long Tail? If so, what are playlists advertising? Does »Barack Obama’s 2019 Summer Playlist«¹¹ propagate the artists represented in it, the individual tracks, the streaming platform or even its curator? And why does even a fruit trading company like Chiquita¹² create playlists? Is a playlist a commodity or a means to market one? Is it only »a way of recommodifying the individual tracks of the disaggregated album,«¹³ as the Spotify Teardown study defines it?

Are playlists therefore, as Jeremy Wade Morris has previously argued, »metacommodities,«¹⁴ meaning that they are »commodities that rewrap individual commodities into a bundle under the assumption that the new whole is greater than the sum of its old parts and another new whole is only a recombination away?«¹⁵ What then distinguishes them from albums? Their flexibility, Morris would argue: »Each playlist puts old commodities into new contexts, offers consumers multiple ways to purchase the same product, and gives users another chance to participate in the process of commodification.«¹⁶ Not the products, but the curatorial act would therefore be the consumed (meta-)commodity.

It is difficult to explain what a playlist actually is because it is subject to even more extreme changes in terms of content and design than other lists before it. Sumanth Gopinath and Jason Stanyek rightfully describe playlists in their essay »Technologies of the Musical Selfie« as »[e]ndlessly fungible and automatically generated, easily stored, widely shared«¹⁷ and refer to their »origins in concert programming and performance setlists, radio and discotheque DJ playlists, the track order of original LPs and reaggregated compilation albums, and end-user-created analogue or digital ›mixtapes‹.«¹⁸

All of this suggests that it is necessary to look at the playlist’s history to gain a deeper understanding of it. It is a history of the dematerialisation of music and its subsequent reaggregation in digital form and thus also of (meta-)commodities whose form changes and whose content is transferred into new contexts through the creation of new ones. This in turn raises socio-cultural questions.

The Production of Popularity: Charts, Radio, and Muzak

Sumanth Gopinath and Jason Stanyek correctly place the playlist in the tradition of concert programmes and so-called setlists, which in turn have their origins in even older practices. However, the repertoires of travelling musicians, troubadours or orally transmitted songs are at best playlists avant la lettre. This is because the history of the playlist only really begins with the letter, or rather the invention of printing press, i.e. the mechanical reproduction of notation.

Starting with the 16th century, compendia of scores were printed and sold for the first time.¹⁹ The beginning of the curation of musical pieces in fixed form thus coincided with the commodification of compositions. While the reproduction of these compositions still required human labour, such collections became mechanically playable as early as the 18th and 19th centuries: the principles of the barrel organ and the pianola, as well as the jukebox invented at the end of the 19th century, were to varying degrees based on the automated playback of pre-selected music.²⁰

As a result of their introduction, the music industry developed a curatorial-recombinatory practice that utilised a variety of media for curation and active playback. It is hardly a coincidence then that, at the same time, lists that catalogue and inventory what is available were becoming more and more relevant and were ultimately having an increasing impact on the consumption and curation of music.

When the US magazine Variety printed the very first charts at the beginning of the 20th century, they were not yet designed as rankings, as became customary with the introduction of the Singles Charts in the British magazine New Musical Express, founded in 1952, or the Billboard 100, which followed three years later. Charts nevertheless always also define what is popular on the one hand and how popularity is constituted in the first place on the other, as Ernest A. Hakanen argues in »Counting down to Number One: The Evolution of the Meaning of Popular Music Charts.«²¹

At the beginning of the 20th century, compilations of scores for domestic use were slowly replaced by the advance of two media at once. At the same time as music was »actually radically rematerialised as a durable commodity, a fixed entity for private consumption,«²² as Martin Scherzinger notes with regard to the emergence of shellac and later vinyl records, the dematerialisation of music allowed the almost infinite recombination of different recordings. This process was also driven by editorially compiled playlists, which in turn are shaped by economic interests.

In radio, commercial interests and cultural curation were not only brought together by a business model that combines the playback of music with the insertion of advertising. The playlists of the programmes also became increasingly influenced by the charts.²³ The fact that the labels, which had meanwhile wrested their supremacy on the music market from the publishers, were actively influencing this became evident in 1960 at the latest, when a law was passed in the USA against so-called »Payola,« the systematic bribery of radio DJs and programme editors.

Payola confirms Jacques Attali’s theory that the culture industry is committed to the »production of demand, not the production of supply.«²⁴ A radio programme always highlights what is popular, just as it proclaims or even produces the popularity of certain songs in the first place. The more often a song is played on the radio, the more deeply it becomes ingrained in the collective memory, thus becoming a hit and selling all the more—which is why it climbs up the charts and is therefore played more often on the radio.²⁵

The entanglement of economic interests and the dissemination of cultural production became all the more apparent in 1934, when the Muzak company began broadcasting its recordings over the radio. The aim of Muzak was, on the one hand, to increase the productivity of factory workers, who were to be manipulated in their work activities with different rhythms and tempos depending on the time of day, and, on the other hand, to provide a friendly background soundtrack from (non-)places of consumption such as department stores. This created a paradoxical situation: although Muzak was supposed to encourage activity, it necessitated passive consumer behaviour.

However, further technological developments in the second half of the 20th century increasingly enabled the public to become curators and recombiners themselves. With the emergence of DJing at the end of the 1960s and 1970s, the technological possibilities of the time were even given expression as a cultural technique. Although these new technologies were not tied to playlists, they were largely responsible for consumers increasingly becoming curators.

At the End of Scarcity: the Walkman Effect, Personalisation, and Dematerialisation

With the widespread introduction of the Walkman by Sony in 1979, another media revolution took place. In his 1987 essay The walkman effect, Shuhei Hosokawa takes a decidedly positive view of the Walkman as a playback device that opens up »eine() Welt des Alleine-Musik-Hörens [a world of alone-listening to music].«²⁶ He praised it as a device of »Verkleinerung [miniaturisation]«²⁷, of »Singularisierung [singularisation]«²⁸, »Autonomie [autonomy]«²⁹ and the »Konstruktion/Dekonstruktion von Bedeutungen [construction/deconstruction of meanings]«³⁰. At the same time, however, he pointed out the associated »Devolution [devolution ]«³¹ in technical-technological terms: »(Der Walkman) repräsentiert eher eine funktionale Reduktion, einen technologischen Rückschritt [(The Walkman) rather represents a functional reduction, a technological regression.].«³²

Hosokawa ignores the development of mixtaping as a cultural technique: individual pieces were recorded onto tape from the radio or from records and later from the compact disc, which arrived on the market in 1983, and then put in an individually determined order, sequenced for example according to thematic aspects or musical-aesthetic parameters. This can be regarded as the first genuinly consumer-curated form of a playlist. With the triumph of the CD and later the CD burner and CD-ROMs in the 1990s, this private playlisting intensified further and further as an evolution of curatorial consumer practices.

This was primarily made possible by the fact that music was becoming dematerialisable as digitisation continued to progress. The MPEG-1 Audio Layer III format was published in 1993 and soon changed the distribution and consumption of music on desktop computers under the somewhat catchier abbreviation MP3. In the meanwhile, the Internet was increasingly becoming the distribution medium for audio content. The multimedia framework QuickTime from Apple and the Real Audio Player from RealNetworks even enabled round-the-clock access to audio files for the first time.

The first Internet radio broadcasts happened as early as 1993. On the 5^th of September 1995, RealNetworks offered the first live broadcast of a baseball game. Audio content progressively became available both on demand and live, enabling a form of personalisation that neither conventional records nor radio could offer on their own terms. The audience became increasingly independent of predetermined playlists and thus also the definitions of popularity provided by the music and distribution industries. From their point of view, everything that had been solid melted into air as if it were plastic like the material that formed the material basis of their product range and thus served as their source of income—it appeared to be melting away before their eyes, while their profit margins did the same.

While mixtaping for the Walkman or the home cassette deck had still been bound to the physical limitations of the cassette tape and was based on the reproducibility of music, it was now possible to copy music pieces (largely) losslessly, while music itself became, at least theoretically, infinitely recombinable in programmes such as the media player Winamp, rolled out in 1997. Whereas music recordings had previously been confined by the logic of scarcity, abundance started to reign supreme. This, in turn, not only had an impact on cultural practices, but also had consequences for the culture industry.

In 1996, the company Liquid Audio began selling music as audio files for the first time, thus becoming the first digital competitor of conventional retail and record stores.³³ The website mp3.com followed shortly afterwards. Its founder, Michael Robertson, offered bands and musicians the opportunity to offer their tracks as free downloads via the portal. Like the later streaming services, mp3.com establishes indirect contact between producers and consumers and thus already hints at the logic of platform capitalism and, more precisely, the functioning of a service like Spotify.³⁴

Shortly afterwards, Napster went online, a service that fanned the flames of the free-for-all mentality on which mp3.com thrived: on the peer-to-peer network, music circulated unregulated between individual users without royalties of any kind flowing to the music industry. After music piracy had for the longest time been tied to physical formats and was accordingly difficult to scale, the dematerialisation and thus copyability of digital music enabled its frictionless, more or less clandestine distribution via the internet between consumers. The result of the emergence of this new sharing economy was that record sales are collapsing.

The iTunes Store, rolled out by Apple in 2003, appeared to be a ray of hope. Through it, music finally became purchasable on a larger scale, either in traditional formats such as the album or EP or »unbundled,«³⁵ i.e. as a single song unit. With the introduction of the iPod two years earlier, the same company had already presented a device whose design was adapted to the new way of listening and allowed a new (self-)organisation of music via the »reaggregation«³⁶ of individual files by the consumer in the form of a playlist.³⁷ This allowed it to be even more dynamic than the classic mixtape. »I wasn’t consuming music so much as curating it […]; I was becoming an organiser, an alphabetiser,«³⁸ wrote Dylan Jones in his 2005 book iPod, Therefore I Am.

But just as the technical-technological devolution of the Walkman brought about new cultural developments, the new technology developed by Apple did not set cultural progress in motion: »I had my own canon, one built on experiences I had when I was back in my teens, when, if I chose to, I would play an album until I liked it, no matter how insubstantial it was,«³⁹ Jones continues, hinting at the fact that while listening behaviour was entering a new age, this was soundtracked by the music of previous eras.

This curatorial shift, which first took place through the emergence of playlists on iTunes as a platform and inside consumers‘ iPods, also became reflected in cultural production. In 2011, the music journalist Simon Reynolds took a pessimistic look back at the preceding years and attested that pop culture was suffering from Retromania, as he called his book on the retro phenomena of the early 21st century. According to Reynolds, global pop culture had lost its temporal and cultural distinctiveness, as exemplified by forms such as the mash-up, two songs mixed together.

»Like the iPod and online music-streaming services, the overall effect is to flatten out all the differences and divisions from music history,«⁴⁰ wrote Reynolds about a process in production that reflects the new habits in reception. From the decade onwards, the latter has been increasingly shaped by (partially) automated programmes.

The Triumph of Metacommodities: Algorithmic Playlists, Datafication, and the Curatorial Turn

While the streaming service Pandora launched in 2005, the service was still largely orientated towards the format of traditional radio. However, one of the central revolutionary features of Winamp—playlisting—had already become »much more thoroughly commodified in the iTunes Store,«⁴¹, as Jeremy Wade Morris notes in Selling Digital Music, Formatting Culture. »In addition to grouping songs by album or artist, as one might find in traditional retail stores, the early iterations of the iTunes Store sorted and sold much of its content through curated playlists.«⁴² These playlists were organised thematically, for example by season or for specific occasions such as Halloween, if they were not being curated by companies.

Analogue to the commercials that have been part of radio since the early days, corporations engaged in marketing by compiling »metacommodities.«⁴³ The sale of music in the form of these metacommodities countered the trend towards more targeted consumption due to increasing unbundling, which meant that the profit margins of platforms, labels and rights holders were lower than in the days of regular phyiscal media sales due to the change in consumer behaviour. It is therefore not surprising that iTunes soon tried to provide more automated purchase incentives through algorithmic control, just as the online marketplace Amazon was perfecting at the time: In 2008, it rolled out a feature called Genius that made recommendations based on consumers‘ iTunes libraries.

However, the foundations for datafication and increasing automation of the internal organisation of download shops and streaming platforms had already been laid before that. It had been Pandora that implemented the Music Genome Project at the beginning of the decade. It »aimed to represent the musical qualitites of tracks in a set quantity of discrete ›genes‹: features such as the gender of the vocalist, the most prominent instrument, the relative tempo, and so on,«⁴⁴ as K. E. Goldschmitt and Nick Seaver describe the efforts that lead to an algorithmically structured organisation of the source material which had been evaluated in this way.

In his essay »Neo-Muzak and the Business of Mood,« Paul Allen Anderson points out that this endeavour bore similarity to the indexing of music as it had been carried out by Muzak decades before.⁴⁵ However, where Muzak focussed on creating specific effects—simulating certain moods and stimulating listeners—this was not yet the case with Pandora. Nevertheless, the service followed the classic »lean back method,«⁴⁶ familiar from radio, as music journalist Mark Hogan writes for Pitchfork: consumers didn’t have to be curators yet.

However, when the Spotify service was launched in 2008, it began as a »lean foward« platform: a »celestial jukebox«⁴⁷ in which listeners could browse through a wide selection of music pieces. The promise was twofold. Not only did the platform allow listeners to discover new music, but it also gave musicians the chance to bypass the traditional distribution channels of the music industry, i.e. to be discovered. Although Spotify had offered playlist functions from the outset⁴⁸, they only came into focus at the end of 2012. The reasons for this are obvious: »Curating millions of songs was a task too large to be left to users.«⁴⁹ The abundance of digital space needed to be reduced in order to make it easier to consume and thus create value.

Hence, a »curatorial turn«⁵⁰ in the company’s strategy took place, as the authors of Spotify Teardown observe. »This meant that Spotify began to transform itself from being a simple distributor of music to the producer of a unique service.«⁵¹ The act of curation itself was commodified and sold. After the music industry had to watch its plastic melt away, it was now solidifying again and presenting itself in the form of a new commodity that was more abstract than anything before. Hence, the music industry was ›rescued‹ after the massive slump in sales figures, i.e. the previous status quo was re-established under new conditions.

Constraints Set the Tone: Playlist Albums, Songwriting, and Spotify-core

The increasing dominance of playlists as metacommodities in Morris‘ sense was also changing cultural production as such. Now that the concept of the album had been »digitally de-ontologised,«⁵² it was itself open to modification. Drake, for example, one of the most streamed musicians ever⁵³,  started releasing albums with a sometimes disproportionate number of songs on them. »This is a symptom of the attention-driven platform economy as well: the churning stomach of the content machine constantly demands new stuff. In such an economy, music that doesn’t take off is dropped once it has outlived its usefulness – either as a brand prop or as playlist-filler,«⁵⁴ writes music journalist Liz Pelly with regard to this trend.

Occasionally, the production and release process already followed the dynamic logic of the playlist: After the initial release of the album The Life of Pablo on the 14^th of February 2016, Kanye West repeatedly changed its track list or even replaced tracks with slightly altered versions of themselves.⁵⁵ Others abandoned the format altogether and shifted to the release of individual tracks in high frequency that were only retrospectively compiled into albums. The album as a cultural form thus reappeared in the dematerialised space in the same form in which it had once materialised in the phyiscal world: as a mere compilation of individual, discrete pieces of music, linked by little more than their authorship.

The album therefore lost relevance, but it did not cease to exist. Since the charts, as a medium for recording and defining popularity, reacted to this unbundling on a grand scale and adapted to the new circumstances by also registering songs that had not officially been declared as singles, paradoxical consequences arose: When Ed Sheeran released ÷ in 2017, all 16 songs on it at once appeared in the top 20 of the British singles charts.⁵⁶ This means that was not anymore the individual songs or albums that were popular, but that—at least according to the new understanding—the impact individual artists have on the platform became the new benchmark for popularity; artists who in turn themselves had become an advertised or rather self-promoting product in this environment.

Songwriting also changed. Many of the bigger hit songs of the second half of the decade were characterised by the inclusion of several musical motifs and sometimes even the chorus within the first thirty seconds as well as occasionally the voices of feature guests. »It’s almost like an executive summary,«⁵⁷ Mark Hogan quotes the songwriter Charlie Harding. The reasons for this were not purely aesthetic, as Hogan emphasises: »In order for a stream to count toward chart tallies and, reportedly, for royalty payouts, a given song must be played for at least 30 seconds.«⁵⁸ Economic constraints came to increasingly set the tone.

The many other phenomena that could be observed in the context of this process included the increased addition of celebrity feature guests to tracks or the release of different versions of a song, but the sound itself was also changing. In The Song Machine. Inside the Hit Factory, John Seabrook describes a gradual shift towards tried-and-tested motifs.⁵⁹ Analogous to the mash-up phenomenon of the previous decade, distinct aesthetic parameters emerged, and they were rooted in the new form of reception moulded by the playlist paradigm.

In this way, the playlist, like the charts before it, came to define what popularity means in economic terms and, consequently, what is popular. The result is »Spotify-core«: »[M]uted, mid-tempo, melancholy pop, a sound that has practically become synonymous with the platform,«⁶⁰ as Pelly defines the non-genre.

A Dual Disappointment: Institutionalised Payola, »Fake Artists,« and the New Gatekeeping

Alan B. Krueger provides an economic explanation for this aesthetic paradigm shift. He speaks of »a tendency for what is popular to become even more popular,«⁶¹ arguing that »playlists that recommend a small set of songs to millions of listeners tend to magnify superstar effects. These features reinforce bandwagon effects and strengthen cumulative advantage benefits for star recording artists.«⁶² Processes such as these were fuelled by labels engaging in »institutionalised payola«⁶³⁶⁴ while at the same time, they run the risk of being replaced by Spotify as a kind of metalabel.⁶⁵ Questions soon arose as to whether the Swedish company not only produced metacommodities in the form of playlists, but also the music itself.

In the summer of 2017, Music Business Worldwide’s Tim Ingham identified at least 50 names for which no information could be found outside of the platform and which, thanks to prominent placement in popular Spotify playlists, were nonetheless able to gather several million streams for themselves.⁶⁶ Very few of the listeners will even have noticed that the tracks in playlists designed to amplify or accompany certain moods or activitys—Ingham lists »Peaceful Piano, Piano In The Background, Deep Focus, Sleep, Ambient Chill and Music For Concentration,«⁶⁷ i.e. classic neo-muzak⁶⁸—came from anonymous people.

However, research carried out shortly afterwards by Dani Deahl and Micah Singleton for The Verge  found that these were »legitimate artists, writers, and session musicians using pseudonyms for a variety of personal and business reasons.«⁶⁹ Nevertheless, theories that Spotify itself disseminated specifically commissioned music via its own platform and placed it in playlists in direct competition with regular artists have not been proven. Still, the case proves how competitive playlist placement is—and how great Spotify’s monopoly on its own platform is considered to be.

The dual promise of discovery and being discovered cannot be realised in this way. Instead, this development contradicts the theory of the »long tail,« as proclaimed by Chris Anderson in his book of the same name in 2006. According to Anderson, big hits would compete with »an infinite number of niche markets, of any size,«⁷⁰ from which, in turn, niche markets in particular would benefit in times of digital abundance. »This has yet to materialize in the music business. Instead, the middle has dropped out of music, as more consumers gravitate to a smaller number of superstars,«⁷¹ comments Krueger dryly in Rockonomics. In this environment, this seems especially true for superstars without faces.

As a result, lesser-known musicians were finding it increasingly difficult to assert themselves in this environment: »In fact, playlists are becoming the new gatekeepers for music«⁷², writes Krueger.⁷³

New Contexts: Discover Weekly, Surveillance Capitalism, and (Self-)Optimisation

In view of the restructuring of reception attitudes, it is often noted that streaming platforms do not, or at least not fully, reflect the contextual affordances provided by physical products such as vinyl records or CDs in the form of artwork, credits, liner notes, lyrics, and other information. Instead, playlists create their very own contexts for their audience.

Following the purchase of the music intelligence company Echo Nest in 2014, Spotify introduced the weekly changing, personalised playlist Discover Weekly shortly afterwards. The feature contributed significantly to the platform’s popularity because it promised both personalisation and convenience—like a mixtape that you don’t have to compile yourself. The recommendation algorithms are fed from data sets that record the behaviour of users far beyond fixed parameters such as age, gender, and others.

The recipients thus perform a »new form of digital labour, extracted by technical interfaces designed for the capture of data,«⁷⁴  as Michael Scherzinger notes. Gigantic feedback loops are created, which in turn trigger new paradigm shifts. The dual promise of discovery inscribed in the title of Discover Weekly should be understood as a form of nudging, as is common in surveillance capitalism.⁷⁵ This new passivity is based upon a type of activity that had little to do with active listening. The algorithms organising playlists are, like anywhere elsewhere in the platform economy, best understood as business interests translated into code.

This explains why Spotify is labelled a »conventional media firm«⁷⁶ in Spotify Teardown: although the company may act as a producer of metacommodities, its actual business lies in a service because it provides third parties with data. »Are you playing the music, or is the music playing you?,«⁷⁷ Seabrook asks rightfully. While Shuhei Hosokawa praised the singularisation and autonomy of the consumer in The walkman effect, this tendency became all the more intensified in streaming. Listeners use a universal interface to access personalised content, and this personalisation leads to a new form of atomisation.

The result of this process is no longer compatible with conventional conceptions of culture since its result necessitates users‘ comparability as they are simultaneously exposed to a homogeneous offering and a heterogeneous experience on a platform. This is also represented by a paradigm shift in playlisting, where the curatorial-recombinatory efforts progressively aim to create and reinforce certain moods or shape activities such as group dinners or even sporting achievements from the background. Liz Pelly notes: »It turns out that playlists have spawned a new type of music listener, one who thinks less about the artist or album they are seeking out, and instead connects with emotions, moods and activities«⁷⁹.

According to Spotify Teardown, a »general self-help ethos and cheerfulness of playlist descriptions«⁷⁸ begins to prevail. Playlists are becoming a (self-)optimisation tool that is used to regulate the emotional balance as if they were pills; an upper in the form of the Beast Mode playlist to start the day, a downer called Ambient Chill to be able to kick back after work. In this context, the playlist becomes not only the most ubiquitous, but literally the most dominant iteration of the order-creating list. »The molding of users into types or taste profiles can itself be seen as an expression of the ›soft biopolitics‹ that […] regulate our lives without us being fully aware of it,«⁸⁰ note the authors of Spotify Teardown.

This implies that platforms such as Spotify do indeed not meet a demand with their playlists, but produce it themselves, much like Jacques Attali observed at the end of the 1970s with regard to the culture industry in general. »Music is prophecy,« wrote the economist and later advisor to François Mitterrand in 1977 in Noise. The Political Economy of Music. »Its styles and economic organization are ahead of the rest of society because it explores, much faster than material reality can, the entire range of possibilities in a given code.«⁸¹ The playlist shapes a cultural change that has social and economic repercussions in other areas of life.

In the course of the history of the playlist, music has been turned into melting plastic; its commodity form was being dissolved and immediately re-aggregated in order to be fitted into a metacommodity that is itself constantly in a state of dynamic change and anticipates the listeners‘ future based on their past. What’s more, it unifies them by homogenising them and the cultural production they consume. In this way, it drives forward processes of neoliberalisation that were initiated long before its creation.

Journalist Cherie Hu, meanwhile, saw the age of the »post-playlist«⁸² as having arrived at the end of 2018 and instead held out the prospect of a form of »artist-centric curation,«⁸³ i.e. an even greater monopolisation of the superstar system. In the meantime, however, apps such as Endel⁸⁴ of course aim to manipulate the listeners‘ mood and represent the increasing automation of the production process, while podcasts are increasingly competing with the music on offer in the streaming environment. However, the playlist as a form and medium continues to enjoy increasing popularity among consumers, whose behaviour it registers and, as some fear, also influences.

At present, the future of the playlist itself remains unwritten. However, what’s certain at the moment is that this text may not reflect recent changes.

1 https://en.wikipedia.org/wiki/Category:Lists_of_music_lists, last accessed on the 6^th of July, 2020.

2 Ibid.

3 See Alan B. Krueger, Rockonomics. What the Music Industry Can Teach Us About Economics (and Our Future), John Murray, London 2019, p. 192.

4 https://support.spotify.com/de/using_spotify/playlists/create-a-playlist/, last accessed on the 6^th of July, 2020.

5 https://de.wikipedia.org/wiki/Wiedergabeliste, last accessed on the 6^th of July, 2020.

6 Ibid.

7 John Seabrook, The Song Machine. Inside the Hit Factory, W.W. Norton, New York 2016, p. 287.

8 Krueger, Rockonomics, p. 192.

9 Vgl. K.E. Goldschmitt and Nick Seaver, »Shaping the Stream: Techniques and Troubles of Algorithmic Recommendation«, in: Nicholas Cook, Monique M. Ingalls and David Trippett, The Cambridge Companion to Music in Digital Culture, Cambridge University Press, Cambridge 2019, pp. 63-81, here p. 65.

10 Chris Anderson, The Long Tail. How Endless Choice is Creating Unlimited Demand, Random House Business Books, London 2006, p. 34.

11 https://open.spotify.com/playlist/3xN6J0LCyVj8k1gVCguWRH, last accessed on the 6^th of July, 2020.

12 https://www.chiquita.de/stickers/chiquita-playlists/, last accessed on the 6^th of July, 2020.

13 Maria Eriksson, Rasmus Fleischer, Anna Johansson, Pelle Snickars and Patrick Vonderau, Spotify Teardown. Inside the Black Box of Streaming Music, The MIT Press, Cambridge MA 2019, p. 117.

14 Jeremy Wade Morris, Selling Digital Music, Formatting Culture, University of California Press, Oakland 2015, p. 162.

15 Ibid.

16 Ibid.

17 Sumanth Gopinath and Jason Stanyek, »Technologies of the Musical Selfie«, in: Nicholas Cook, Monique M. Ingalls and David Trippett, The Cambridge Companion to Music in Digital Culture, Cambridge University Press, Cambridge 2019, pp. 89-118, here p. 97.

18 Ibid.

19 See Jacques Attali, Noise. The Political Economy of Music, trans. Brian Massumi, University of Minnesota Press, Minneapolis/London 1985, pp. 51-55. See also Ernest A. Hakanen, »Counting down to Number One: The Evolution of the Meaning of Popular Music Charts«, Popular Music, Volume 17, pp. 95-111, here p. 102.

20 See Martin Scherzinger, »Toward a History of Digital Music: New Technologies, Business Practices and Intellectual Property Regimes«, in: Nicholas Cook, Monique M. Ingalls and David Trippett, The Cambridge Companion to Music in Digital Culture, Cambridge University Press, Cambridge 2019, pp. 33-57, here p. 35.

21 See Hakanen, »Counting down to Number One«, p. 97.

22 Scherzinger, »Toward a History of Digital Music«, p. 39.

23 Hakanen, »Counting down to Number One«, p. 97.

24 Attali, Noise, p. 103.

25 In the (admittedly exceptional) case of Mariah Carey’s »All I Want for Christmas« or Wham!’s »Last Christmas,« such a virtuous circle can occur regularly ad infinitum or at least ad nauseam.

26 Shuhei Hosokawa, Der Walkman-Effekt, trans. Birger Ollrogge, Merve, Berlin 1987, p. 12.

27 Ibid., p. 16.

28 Ibid., p. 17.

29 Ibid., p. 18.

30 Ibid., p. 20.

31 Ibid., p. 14.

32 Ibid.

33 See John Alderman, Sonic Boom. Napster, P2P and the New Pioneers of Music, Fourth Estate, London 2002, pp. 44-46.

34 »Platforms, in sum, are a new type of firm; they are characterised by providing the infrastructure, to intermediate between different user groups, by displaying monopoly tendencies driven by network effects, by employing cross-subsidisation to draw in different user groups, and by having a designed core architecture that governs the interaction possibilities.« Nick Srnicek, Platform Capitalism, Polity, Cambrdige/Malden 2017, p. 48.

35 Krueger, Rockonomics, p. 197.

36 Morris, Selling Digital Music, p. 159.

37 »Napster […] had effectively laid the groundwork for Apple’s rise to market dominance, forging the way toward an efficient and interactive new model for music listening,« writes Scherzinger. Scherzinger, »Toward a History of Digital Music« p. 44.

38 Dylan Jones, iPod, Therefore I Am, Phoenix, London 2005, p. 19.

39 Ibid., p. 21.

40 Simon Reynolds, Retromania. Pop Culture’s Addiction to Its Own Past, Faber and Faber, New York 2011, p. 360.

41 Morris, Selling Digital Music, p. 159.

42 Ibid.

43 Ibid., p. 160.

44 Goldschmitt and Seaver, »Shaping the Stream«, p. 70.

45 Paul Allen Anderson, »Neo-Muzak and the Business of Mood«, Critical Inquiry, Volume 41, pp. 811-840, here p. 823.

46 https://pitchfork.com/features/article/9686-up-next-how-playlists-are-curating-the-future-of-music/, last accessed on the 6^th of July, 2020.

47 The term was already coined in 1994 by Paul Goldstein and has successively applied to Spotify and other streaming services. See also https://freedom-to-tinker.com/2012/09/18/is-spotify-the-celestial-jukebox-for-music/, last accessed on the 6^th of July, 2020.

48 See Eriksson, Fleischer, Johansson, Snickars and Vonderau, Spotify Teardown, p. 117.

49 Ibid., S. 61.

50 Ibid.

51 Ibid.

52 Scherzinger, »Toward a History of Digital Music«, p. 280.

53 See also https://www.washingtonpost.com/arts-entertainment/2019/12/03/drake-was-spotifys-most-streamed-artist-decade-what-does-that-actually-mean/, last accessed on the 6^th of July, 2020.

54 https://thebaffler.com/downstream/streambait-pop-pelly, last accessed on the 6^th of July, 2020. Juli 2020.

55 These were obviously documented as lists. See also https://www.xxlmag.com/kanye-west-the-life-of-pablo-changes/, last accessed on the 6^th of July, 2020.

56 https://www.theguardian.com/music/musicblog/2017/mar/10/ed-sheeran-has-16-songs-in-the-top-20-and-its-a-sign-of-how-sick-the-charts-are, last accessed on the 6^th of July, 2020.

57 https://pitchfork.com/features/article/uncovering-how-streaming-is-changing-the-sound-of-pop/, last accessed on the 6^th of July, 2020.

58 Ibid.

59 A songwriter such as Dr. Luke, sued several for plagiarism, follows what Seabrook calls the »Rule of Three.« According to this logic, listeners are only able to judge a song after having heard it three times. However, if a piece is reminiscent of a famous song, this process is significantly shortened. Seabrook, The Song Machine, p. 248. See also Krueger, Rockonomics, p. 10.

60 Pelly, Streambait Pop.

61 Krueger, Rockonomics, p. 86.

62 Ibid., p. 193.

63 This is how Fredric Dannen called it in his chronicle of the US-American music industry since the 1950s. Fredric Dannen, Hit Man. Makler der Macht und das schnelle Geld im Musikgeschäft, trans. Peter Robert, Zweitausendeins, Frankfurt am Main 1998, p. 25. In fact, payola is not illegal in the streaming environment.

64 For the release of Drake’s Scorpion, songs from the album were prominently included in even completely unrelated lists such as Best of British (although various tracks are inspired by British grime music, Drake is a Canadian citizen) thanks to deals between Spotify and the artist involved. Apple even advertises it in conjunction with the Siri voice assistant app. See Krueger, Rockonomics, p. 182.

65 »Another development to watch is that streaming companies can attempt to compete with traditional record labels and publishers, much as Netflix has done with movie studios. Spotify has encouraged artists to post music on Spotify directly, without a label […].« Krueger, Rockonomics, p. 202. Pelly puts it more drastically: »No matter how you look at it, it’s clear that Spotify is trying to replace labels.« Pelly, Muzak.

66 https://www.musicbusinessworldwide.com/spotify-denies-its-playlisting-fake-artists-so-why-are-all-these-fake-artists-on-its-playlists/, last accessed on the 6^th of July, 2020.

67 Ibid.

68 See Anderson, Neo-Muzak and the Business of Mood.

69 https://www.theverge.com/2017/7/12/15961416/spotify-fake-artist-controversy-mystery-tracks, last accessed on the 6^th of July, 2020.

70 Anderson, The Long Tail, p. 5.

71 Krueger, Rockonomics, p. 13.

72 Ibid., p. 193.

73 What’s more, this facilitates social imbalances, as Pelly points out her text »Discover Weakly.« In a self-experiment, the journalist created a new Spotify profile and relied purely on the platform’s recommendation function, only to have her data profile analysed for gender distribution by a feature jointly offered by Spotify in collaboration with an alcohol brand: »Indeed, when I signed into the Smirnoff Equalizer last month, in the midst of my listening experiment, the app concluded that I had been listening to 100 percent male artists. What actually happened: I had merely been listening to what Spotify told me to listen to by way of its playlists.« https://thebaffler.com/latest/discover-weakly-pelly, last accessed on the 6^th of July, 2020.

74 Scherzinger, »Toward a History of Digital Music«, p. 48.

75 See Shoshana Zuboff, The Age of Surveillance Capitalism. The Fight for the Future at the New Frontier of Power, Profile Books, London 2019, pp. 293-309 as well as Anna-Verena Nosthoff and Felix Maschewski, Die Gesellschaft der Wearables. Digitale Verführung und soziale Kontrolle, Nicolai Publishing & Intelligence GmbH, Berlin 2019, pp. 30-34.

76 Eriksson, Fleischer, Johansson, Snickars and Vonderau, Spotify Teardown, p. 164.

77 Seabrook, The Song Machine, p. 289.

78 Eriksson, Fleischer, Johansson, Snickars and Vonderau, Spotify Teardown, p. 125.

79 https://thebaffler.com/salvos/the-problem-with-muzak-pelly, last accessed on the 6^th of July, 2020.

80 Eriksson, Fleischer, Johansson, Snickars und Vonderau, Spotify Teardown, pp. 136/137.

81 Attali, Noise, p. 11.

82 https://www.getrevue.co/profile/cheriehu42/issues/our-new-post-playlist-reality-138493, last accessed on the 6^th of July, 2020.

83 https://www.patreon.com/posts/celestial-era-is-32326028, last accessed on the 6^th of July, 2020.

84 Vgl. https://spex.de/endel-die-welt-als-wohlfuehlblase/, last accessed on the 6^th of July, 2020.

85 Vgl. u.a. https://www.horizont.at/digital/news/spotify-millionen-schlacht-um-podcasts-81420, last accessed on the 6^th of July, 2020.