Endless Summer
Transcultural Geographies of City Pop Art
Image credit: Adeel Zafar, Taxiway: Hiroshi Nagai Reimagined (2020)1
And Suddenly There Was City Pop: The Global Dimensions of the Retrograde Formation of a Japanese Popular Music Genre
University of Amsterdam, 19-21 March 2025
I. Double Fantasy
My intention when I first proposed this paper was to focus not on the music of city pop but on the artwork of Hiroshi Nagai that has become its visual counterpart. That was before I actually did my homework and discovered that the work has already been substantially done for me—notably by our dear colleague Toshiyuki Ohwada, whose excellent article “City Pop’s America” is among the three essays included in the catalog to the exhibition that ran at the Japan Foundation Gallery in Sydney, Australia, from September 2020 to January 2021, titled Hiroshi Nagai: Paintings for Music, concurrently with a second exhibition, City Pop: Inspired Nostalgia. Since the full catalog to the Nagai exhibition is available online, as well as an extensive discussion of Nagai’s work by Professor Ohwada and Mark McNeill, I can only refer colleagues to these materials for the background that I originally planned to provide. Since the YouTube algorithm launched the viral success of Mariya Takeuchi’s 1984 hit “Plastic Love” (St. Michel, 2018), and the subsequent release of Mark McNeill’s expertly-curated Pacific Breeze compilation for Light in the Attic Records (2019), Nagai’s work has become an increasing focus of attention as the visual representative of city pop. Discogs lists him as having illustration credits for some 144 vinyl and CD releases. As with Korea’s rapid (not to mention lucrative) international institutionalization of K-pop, the Japan Foundation has been quick to seize an opportunity to flex some Japanese soft power of its own. In Spring 2023 the Bunkamura gallery in Shibuya hosted a homegrown retrospective, 永井博 The Journey Never Ends. Nagai’s images are ubiquitous on YouTube’s infinity pool of city pop playlists and fan videos, on Instagram, Pinterest, wallpapers, animated gifs, a subreddit, and they have even inspired original music tracks.
For a sampler of Nagai’s album covers, see Anton Spice, “The sun-drenched Americana of Japanese artist Hiroshi Nagai” (The Vinyl Factory, 24 April 2019).
Given all this, I will not be reprising here Nagai’s work and his involvement in both the origins and afterlife of city pop, from the 1979 publication of his book A Long Vacation and its featuring on Eiichi Ohtaki’s city pop classic of the same name (1981) to the Light in the Attic series and beyond. Nor will I be framing Nagai’s Californian dreamscapes in relation to the thematics of pastiche and exotica elaborated in Shuhei Hosokawa’s concept of “soy sauce music” (even though I love it) and Haruomi Hosono’s Tropical Dandy alter ego. The abundance of existing documentation on Nagai’s work at least clears some space for us to check out some of the lesser-known islands in the archipelago.
What I want to do here is to pick up where the existing discussion leaves off and consider some of its more interesting threads—for example, the observation by Yamato Shiine (editor of Popeye magazine), cited by Toshiyuki Ohwada, about how the lifestyle sports of surfing, skate boarding, and hang gliding popularized in 1970s California were about weightlessness, a sense of floating, and a feeling of equilibrium (Shiine 2010, cited in Ohwada 2020), prompting the following point from Ohwada:
This abstract sense of floating can be seen in Hiroshi Nagai’s work. Its sense of weightlessness is probably derived from the artist’s origins in surrealism in addition to a shared sensibility of American west coast sports (Ohwada 2020: 11).
What interests me here (in addition to the impact of Surrealism on Nagai’s work) is that this is a point about aesthetics, since aesthetics is what I want to focus on in what follows, albeit in a different sense from how that term has been tradition understood in art history.
While the bulk of existing scholarship on city pop to date has been historical, in terms of contextualizing the emergence of the style itself in relation to larger economic forces—basically the bubble economy of the 1980s, while it lasted—and their socio-cultural shifts, as well as the more recent history of “city pop” itself as a retroactive form of nostalgia for a future that never materialized, my orientation here is more resolutely presentist. I am interested in the here and now of city pop as a dynamic force of contemporary cultural content. In particular, I am interested not only in the economic and socio-cultural forces that generated it, but the new media and cultural practices being generated from it. Mainly, though, I am interested in city pop as a contemporary example of what has come to be known as an Internet aesthetic.
II. Summer Feeling
To elaborate on this, some methodological observations are first necessary. From a disciplinary standpoint, city pop has been primarily approached to date as music—that is, within the discursive formation of popular music studies, specifically the history of Japanese popular music, and global/transnational music production, reception, and consumption. Within this formation, while Hiroshi Nagai’s illustrations are acknowledged to be a key visual component of city pop aesthetics, they are still conceived as somewhat auxiliary to it; primarily, it is assumed, “city pop” refers to a style of music, albeit one with a strong visual counterpart, but primarily music nonetheless. Much the same could be said of the visual components of other Japanese popular music genres, notably Shibuya-kei and, of course, Visual-kei. But it’s precisely this music-centric conception of city pop that I want to problematize here. One of the limitations of the popular-music-studies approach, which has already begun to emerge in its difficulty in coming to grips with new aesthetic forms such as vaporwave or lo-fi hip hop, is its inadequacy in conceptualizing the role of social media in their production and circulation. While they are obviously referenced as having “originated” on social platforms and/or “the internet”, it’s still assumed that what the primary object of attention are and should be musical practices and aesthetics. While city pop is frequently mentioned in connection with Internet aesthetics like vaporwave or future funk (“Introduction”: 4), such references are typically made only in passing before returning to the music itself, its artists, and their history. YouTube is considered as no more than a platform for the music rather than an active agent of cultural production. Against this framework, I propose that global cultural forms such as city pop, lo-fi hip hop, or vaporwave should be approached not as primarily musical styles alone but as aesthetics, in which music is only one element with equivalent status to artwork, video, fashion, memes, and other forms of digital content.
So what exactly is an “aesthetic” in this context, and what does it mean to reframe city pop as an aesthetic, rather than what it rather obviously appears to be: a historical genre of Japanese popular music? In what is to my knowledge the only academic source to date to have explored the concept of the aesthetic in theoretical terms, Guilherme Giolo and Michaël Berghman (2023) suggest that traditional conceptual frameworks such as genre, lifestyle, or subculture are no longer adequate to understanding the contemporary modalities of cultural identity and creativity. Instead, they explore the ubiquitous popular concept of the “aesthetic” as the key component of a new analytical framework. What does an “aesthetic” in this contemporary sense consist of? We might start by examining the entry on “City Pop” on the Aesthetics Wiki, a vast online catalogue of aesthetic styles from “Algorave” to “Zombie Apocalyse”, “Dark Academia” to “Visual-kei”, “Cottagecore” to “Witch House”. Like many of the other entries about aesthetics that the reader is already acquainted with, however, the city pop entry, while a useful introduction, brings few surprises, opening with a bland definition of it as “a subgenre of Pop music from Japan”. The remainder of the entry is a walkthrough of everything that everyone at this conference already knows, with the obligatory references to 1970s American AOR, YMO, and the California Dreaming of Hiroshi Nagai. What the entry doesn’t get us much closer to, however, is understanding why city pop can be considered not just as a J-pop subgenre but an aesthetic. For that we need to go back to Giolo and Berghman.
One of the defining components of what Giolo and Berghman call Internet aesthetics is their association with certain kinds of experience, and the corresponding feelings evoked by the memory of that experience: the experience of living in the world of 1950s America, for example, or the world of Harry Potter; the experience of living in a dystopian future; or, one might add, the experience of living on the American West Coast in the 1970s, or in Tokyo at the height of the bubble economy. While music is a key component of such experiences, it is only one modality within a larger cultural world. Imaginatively speaking, aesthetics recall the fantasy worlds of videogames and LARPing:
Experiences (and, likely, their importance for Internet aesthetics) seem to be even more evident in the titles of playlists associated with different aesthetics: “You are studying in a haunted library with ghosts” (dark academia playlist), “You found the entrance to a secret garden”, “You are falling for the protagonist in a fantasy novel” (light royaltycore), or “A playlist for old money living in the French countryside” (light academia). Even clearer here, multiple objects (songs) are grouped and given symbolic meaning under a new interpretive key. This key is experience-centered. Thus, it seems acceptable to assume that experience is also the goal of an Internet aesthetic: to become someone different, to feel as if living in a different time (2023) [Emphasis mine]
The dominant musical modality of aesthetics is the playlist, and playlists can be thought of as scripts or catalysts for certain kinds of experience: the dark cinematic vibe of film noir in jazz; chillout-room electronica; the moody self-absorption of shoegaze; the entire “After Hours” DJ session collection. Although originating in the album-based formats of the 1970s and 1980s, city pop today is something that we experience as much as just listen to, something that conjures up a particular structure of feeling (as Raymond Williams used to call it); and the primary way we experience it is through the playlist, whether in the form of the compilation release (Pacific Breeze) or the endless-summer playlists of YouTube and Spotify. With the advent of the playlist, all music became background music—music for something, that creates the ambience for certain kinds of experience: studying, barbecueing, driving at night, lovemaking, being depressed after a breakup. Music has long had this function, of course; its origins go back to the so-called “mood music” of the 1950s, when stereophonic recording began to turn it into an environment: Music For Lovers; Soothing Sounds for Baby; and several decades later, Music For Airports. Since the 1990s, mood music has been replaced by the concept of vibe, but both are arguably about music as a catalyst for shared experience (or the imagination of such experience) and the structures of feeling associated with them. It’s no coincidence that the current generation of city pop fans around the globe are too young to have lived through the decades to which it provided the musical soundtrack, since anemoia—–nostalgia for something one never experienced directly—–is one of the key constituents of Internet aesthetics. To listen to an album today such as Tatsurō Yamashita’s appropriately-titled For You (1982), most of us can only dream of another, better time and place, when Japan was still surfing the economic wave and the prevailing mood was for glamour and partying. A decade later, after the bubble had burst, while Yasuharu Konishi’s Readymade Records continued to party like it was (not yet) 1999, the vibe of Shibuya-kei led by Crue-el and Trattoria Records was cooler, more circumspect, more indie—Music For The Morning After.
Giolo and Berghman also emphasize the affective dimension of Internet aesthetics, and their interviewees frequently reference feelings when describing them:
“The aesthetics that I like really depend on my mood of the day. On Sundays, when it’s sunny, I really enjoy aesthetic images that give me this summer feeling” Jane (Dutch, prefers “cottagecore”).
Jane, who strongly connects “cottagecore” to morning walks, went on to explain that in those moments the scenarios and imaginaries of this Internet aesthetic “flow” into her activity, representing “the things I would like at that moment. They give me a warming feeling” (2023). [Emphasis mine]
By now we are (hopefully) getting closer to understanding what it means to describe city pop as an aesthetic rather than just a musical genre. Like vaporwave, like future funk, like mallsoft, like lo-fi hip hop, city pop conjures up a particular kind of cultural-historical aesthetic experience and a corresponding affective structure, via playlists, imagery, and other elements. As has been much discussed in the case of vaporwave, the dominant temperamental register is nostalgic and melancholic, invoking a bygone era of economic optimism and technological utopianism rendered only more poignant by the harsh realities of the dystopian present.
III. Infinity Pool
If the preceding discussion has led you to consider reframing city pop as an aesthetic, let me conclude with an example of how that aesthetic is continuing to mutate into new forms of creative expression. One of the characteristics of an aesthetic is that it is defined by–and can be encoded as–a set of audiovisual stylistic parameters that constitute its aesthetic fingerprint. As many recent studies show, the algorithms of streaming platforms such as Spotify are based on the systematic encoding of such styles, often referred to as a kind of musical DNA. Spotify, for example, has three official city pop playlists, respectively for the 1970s, 1980s, and 1990s. This algorithmic systematization of popular music has had a number of unexpected consequences, one of which has been the much-discussed (and much-maligned) atomization of music genres: the explosive proliferation of algorithmically-defined styles based solely on audience listening data. As the website Every Noise At Once attests, this musical ecosystem continues to mutate and expand at an almost alarming rate.
The other, even more frequently denounced consequence of the parameterization of musical styles has been automation: the algorithmic generation of content using models trained on large datasets of historical artistic styles, based on a textual prompt input by the user. While far from alone in this respect, the music of city pop is among the styles in question. A simple search on YouTube using terms like “city pop ai” quickly produces a large number of playlists of AI city pop music.
As Claude.ai calmy explained to me but presumably everyone here already knows,
There is indeed a growing amount of AI-generated city pop music on YouTube. City pop’s nostalgic 80s Japanese aesthetic, with its distinctive synthesizer sounds, funk/disco influences, and cityscapes has made it a popular genre for AI music generation.
Many creators are using AI tools like Suno, Udio, and other music generation platforms to create city pop tracks that emulate the style of artists like Mariya Takeuchi, Tatsuro Yamashita, and other city pop icons. These AI-generated tracks often feature the genre’s signature elements - jazzy chord progressions, funky bass lines, and smooth saxophone solos.
The popularity of city pop’s aesthetic combined with the increasing capabilities of AI music generation has led to numerous channels dedicated to sharing these AI-created city pop tracks, often paired with anime-style visuals or cityscapes.
Ending on a more reassuring note, it added that
while there’s definitely AI-generated city pop on YouTube, it still represents just a portion of the city pop content there, with many channels still focused on sharing original city pop tracks from the 70s-80s Japanese music scene or human-created new music in that style.
There are of course many ways to do this, depending on your level of Python coding skills. On Udio, the following prompt generated a wallpaper sample—inevitably titled “Neon Dreams”—that was as generic as you would expect.
Japanese city pop, 1970s era, funky beat, complex electric piano jazz chords, synthesizer flourishes, evoking nostalgia for American album-oriented rock and California lifestyles.
Today’s diffusion models are by no means limited to musical outputs, moreover, and along with generative playlists and music, city pop’s visual components have also begun to be automated. One of the cornerstones of today’s generative media aesthetics is restylization, the application of historical styles of art, music, cinema, or other media to any kind of audiovisual content. A frame-grab from a Studio Ghibli film, for example, can be algorithmically mapped onto the movements of a dancer in a live action video, so it appears to be an animated character that is dancing; but the source image could just as easily be a medieval painting. Such imagery has become routine on TikTok, while a rising tide of open-source models uploaded to websites such as CivitAI and Hugging Face enable users to generate audiovisual content more customizable than better-known sites like Midjourney or Runway. On CivitAI, a simple search for “Hiroshi Nagai” produces numerous city pop style models trained on datasets of images by the painter. These images were generated using one such model; they are not copies of specific works, but none of them are the work of Nagai himself.
As the text on the model page explains:
This LoRA brings the nostalgic charm of Hiroshi Nagai’s signature pop art style to Flux Dev, an open-source image generation LLM. Known for his vibrant depictions of 1970s-80s urban and beach life in Japan, Nagai’s art captures a dreamy yet minimalistic aesthetic—bold colors, crisp lines, and an atmosphere of leisure and tranquility.
With this LoRA, you can generate stunning city pop-inspired illustrations featuring neon-lit cityscapes, serene coastlines, palm-lined streets, and retro-modern architecture. Whether you’re creating vintage-inspired album covers, dreamy summer scenes, or classic city pop aesthetics, this model helps you bring that unmistakable Nagai vibe to your artwork.
This LoRA was developed just one month after the initial release of Flux Dev, contributing early to its expanding creative possibilities. And still, in 2025, Flux Dev remains the leading open-source image generation LLM. Try it out and share your creations! 🚀
Created for Flux Dev by Burç Buluklu.
Since images can also be used as source inputs for time-based generative content, and animation is one of the fasting-growing areas of generative AI, it seems likely that we may soon see generative city pop music videos or animated short movies that that bring Hiroshi Nagai’s images to life, as the film Loving Vincent did with Van Gogh’s iconic paintings. I am neither condoning nor condemning these practices; in engaging with them, I seek only to better understand them by direct experience of their mechanisms.
While for many fans of city pop such developments are likely to be seen as the moral equivalent of dancing with the devil, if you would permit me to take on the role of devil’s advocate for a moment, one might ask what is ultimately so terrible about all this. Is it not simply an automated version of what city pop itself has been doing all along—stylistic pastiche—except that in this case the object of that pastiche is the aesthetic of city pop itself? Is it not just a technologically-enhanced form of sampling, another musical technology with an equally troubled history, but one that is also part of city pop’s own history? Whatever one may think of it, this is the cultural reality of city pop today. For me, the most useful way of thinking about these strange new mutations of city pop is the musical concept of harmonics, the overtone sounds that resonate at higher multiples of a given frequency, known as the fundamental (if this sounds technical, think of Steve Howe’s guitar at the beginning of Yes’s famous song “Roundabout”). From this perspective, rather than simply dismissing images generated from Nagai’s we might think of them as resonating with them at different frequencies, like ripples spreading across a pond, or the shimmering patterns of sunlight on the surface of a swimming pool. Given the prominence of swimming pools in Nagai’s own work, it seems a suitably appropriate metaphor.
City pop generative music and images raise highly-charged questions about the ethics, legality, and aesthetics of media practices in the algorithmic age: about copyright and the concept of digital rights itself; about the place of the human in artistic creativity; about authorship, and about the status of generative art. Whatever our answers to such questions, new creative practices continue to emerge and will continue to do so. The summer of city pop may indeed turn out to be endless, but not in the way anyone back in the heady days of the bubble economy could ever have imagined. As we look back on city pop and its legacy, already a quarter of the way into a new century, we might also turn our gaze in the other direction, and contemplate its possible futures.
Image generated by the author
16 March 2025
Model: FLUX Dev.1
LoRA: Hiroshi Nagai Style Pop Art
Footnotes
Zafar’s project is a 3D animation and video loop adapted from Hiroshi Nagai’s cover art for the 7-inch remix album Record Players 2 by Yakenohara x Chabe (Jetset, 2016).↩︎