Playlist for Prefab Houses: YouTube and the Revitalization of Japanese Ambient Music
“Thank science for algorithms because they brought me here, to my destiny.” - YouTube user “Becky” on Hiroshi Yoshimura’s Soundscape 1 surround (1986)
Liz Pelly’s recent article for The Baffler, “The Problem with Muzak,” bemoans music journalism’s embrace of Spotify. Algorithmically fueled, mood-based playlists such as “Ambient Chill,” she argues, are nothing more than “emotional wallpaper” for the distracted, disengaged masses. As both a writer and record collector, I wholly share Pelly’s concern—the automation of taste would mean the end of criticism. Yet I can’t help but wonder at the Deep Learning Architectures of algorithms, especially that of YouTube—Spotify’s comparably ill-fitted, music streaming cousin. Buttressed by the blogs, playlists, and Dropboxes of zealot crate diggers across the globe, its algorithm has become startlingly intuitive in the past few years. Serving as a favored platform for out-of-print and obscure rarities, YouTube has helped stimulate a revival in the most surprising of genres: 1980s Japanese electronic and ambient music.
The international infatuation with Japan’s bubble-era gems was at first limited to a handful of enthusiasts—DJs in the US and Europe taken with the curatorial magic of scholarly record shop owners like Dubby (Om-Ondas), Chee Shimizu (Organic Music), and Norio Sato (Rare Groove). The buzz congealed online in early 2010, largely through “夕方の影 (Evening Shadows)”—a SoundCloud mix by Music from Memory’s Jamie Tiller. Featuring the lilting synth and breathy sax of categorically transcendent groups like Mkwaju Ensemble and Mariah, the stream was quickly littered with ID requests (SoundCloud mixes rarely feature track lists). A few months later, Spencer Doran—of the electronic duo Visible Cloaks—built on the fever with his own mix for Root Strata, “Fairlights, Mallets and Bamboo: Fourth-world Japan, years 1980 – 1986.” Marching and meditative, new wave and old world, the playlist proved infectious.
Unlike Tiller’s “Evening Shadows,” Doran’s mix included a track list. Curious listeners began to research the featured artists, digging up their long out-of-print albums and sharing them online. “I can imagine that YouTube was very useful,” Chee Shimizu told me by email. Following a record store pop-up at London’s LN-CC, as well as his own mixes of largely unknown Japanese ambient tracks, Shimizu began getting “a lot of inquiries from overseas people about stock conditions of those records.” Throughout the early 2010s, curiosity spread across his DJ friends in Europe, the dance floor, online music blogs, and, of course, down the algorithmic rabbit hole itself—YouTube.
Bouncing between Doran and Shimizu on either side of the Pacific, a cross-cultural connection was quickly made between musicians, playlists, and platforms. In the past six or seven years, anyone looking up western composers like Harold Budd, Brian Eno, or Klaus Schulze likely earned a YouTube autoplay peppered with Japanese electronic and new age albums. As Root Strata’s Rootmixes begot listeners begot YouTube streams, the interest in these albums became enthused. Tracking rising popularity, quantifiable through play counts, labels eventually caught on. Since 2015, I’ve watched my YouTube autoplay queue up for vinyl reissue—from YMO member Haruomi Hosono’s Philharmony (1982) to Yasuaki Shimizu’s solo albums Kakashi (1982) and IQ 179 (1981). One of the most notable successes has been Palto Flats’ 2015 reissue of Mariah’s luscious Utakata No Hibi. The 1983 masterpiece of Fourth World fusion sold out multiple pressings, frequented year-end lists and, most surprisingly, now plays regularly at the bar downstairs.
Such cascading enthusiasm is symptomatic of our times. Like Facebook, YouTube’s algorithm naturally inflames subcultures—conspiracy theorists and vinyl enthusiasts alike. YouTube engineers claim their Deep Learning Algorithm, modeled after our brain’s neural network, is one of the “largest-scale and most sophisticated” recommendation systems. As the algorithm evolves, even more nuanced and intelligent connections between users’ search histories and tastes are formed, amplifying niche audiences—for better or worse.
The centerpiece of YouTube’s recent algorithmic influence is 1983’s Through the Looking Glass, the debut solo album of Mkwaju Ensemble’s Midori Takada. A treasured rarity within music circles for decades, this masterwork of bubble-era Japanese minimalism was largely inaccessible until 2013, when musician Maxwell August Croy uploaded it to the Root Strata blog. A few days later, Takada was featured in another RootMix by Doran, “Music Interiors,” a compilation of 1980s “Japanese new-age/ambient/minimalist music.” Like the Fourth World collection before it, listeners fleshed out the track list on YouTube. Blogger Jackamo Brown’s early upload of the album netted nearly 2 million plays before it was taken down in 2017, reincarnated as a vinyl reissue through Palto Flats and WRWTFWW Records. Incredibly, this was the number one new release on Discogs’s mid-2017 report—a buxom statistic for the typically fringe genre.
Much has been written about YouTube’s influence in the success of this album. However, it was the musicians, DJs, and obsessives who first trained our eyes on Through The Looking Glass that also, indirectly, trained the algorithm. This is exactly why listeners have been so supportive. Engaged in the community, they’ve driven vinyl reissues and energized the careers of the musicians they admire (both Takada and Yasuaki Shimizu are touring Europe this summer). This enthusiasm has unearthed and reinstated the catalogs of legendary Japanese labels like Better Days and P.S.F., and even encouraged the creation of new labels. In 2017, Doran and Croy joined to launch the new imprint Empire of Signs. Named after Roland Barthes’s book of ruminations on a satori-spiked Japan, the label’s first release had been long in the works: a reissue of 1982’s Music for Nine Postcards—the first album by late ambient pioneer, Hiroshi Yoshimura.
True to its name, Postcards is all intimacy and distance, warmth and space. In the liner notes, Yoshimura shares how he made music while gazing out his window, transcribing “sound fragments” onto postcards that he later sent to friends. Released through Satoshi Ashikawa’s Wave Notation series, Postcards is an early example of Yoshimura and Ashikawa’s “environmental” music (kankyō ongaku): “a more conscious attitude toward the sounds—other than music—that we listen to.” Inspired largely by Brian Eno, Erik Satie’s “furniture” music, and acoustic ecologist R. Murray Schafer, environmental music was composed as an audile salve, an infusion of nature meant to combat the anxiety of urban life in 1980s Japan. This is best demonstrated in Green (1986), Yoshimura’s verdant soundscape of falling water and flittering birds. Laid out under stratospheres of mille-feuille synth and smelling of wet grass, its three-dimensionality epitomizes the musician’s devotion to “sound design”—music as carefully considered as a museum’s architecture. Like many of his contemporaries, Yoshimura also applied this idea to his brand collaborations: 1984’s A.I.R. was figuratively and literally fragrant with Shiseido’s forest-scented perfume, while Soundscape 1: Surround (1986) came with the purchase of a prefab house.
Taken with the view from a window of the Hara Museum of Contemporary Art, Yoshimura insisted his music belonged there. Where does it belong now? Readapted to the environment of the Internet, moving through our Deep Learning Architectures and onto LP shelves, this music seems to have found a complementary space—a habitat of web and wax. Indeed, later this summer, Light in the Attic Records will be releasing Kankyō Ongaku: Japanese Ambient, Environmental & New Age Music 1980-1990.
In the last chapter of The Empire of Signs, Barthes concludes that “it is no longer the great continuous wall which defines space, but the very abstraction of the fragments of view (of the ‘views’) which frame me.” Thinking of Yoshimura’s windows, I open a new one: is it possible to scale up a healthy relationship between algorithm and curation, between the architecture and the art? From this view, at least, I’m feeling hopeful.