Philosophy dabbler. Anime ace.
Introduction
“certain aspects of Japanese cartoons […] turn me off, and surprisingly these elements remain consistent throughout all of the anime I’ve seen, whether it’s a cheaply-produced OAV or a classy Miyazaki production. Namely, it is the unappealing and cold nature of their character designs, and the general lack of dynamics and complexity in their personality animation. The death of animation is if you don’t find the characters believable because subsequently the value and effectiveness of the stories those characters are telling is diminished.” —Amid Amidi2
From Michel Ciment’s first use of the term “cinema of slowness” in 2003 to the widespread, heated debate Nick James set off in the April 2010 issue of Sight and Sound, slow cinema became a major topic among film scholars and critics in the early twenty-first century.3 As the discussion developed, critics and theorists consciously traced its origins to modern cinema, from neorealism and the French New Wave onward. Though only recently the focus of sustained discussion, slow cinema had already taken root sixty or seventy years earlier in attempts to break with classical Hollywood narrative conventions. It has a long tradition behind it.
For animation, however, the situation is quite different. Cinema reproduces movement through an automatism that removes human intervention; animation, by contrast, fundamentally needs to “create” movement, to “create” life—as the meaning “to give life” of its Latin root animus suggests. It is not the image-making machine that generates movement. The transformation of stillness—drawing—into movement—animation—makes this an art concerned with movement from the outset. In practice, animation has therefore always been an art form with nothing to do with “stillness” or “slowness.” Disney’s twelve principles of animation teach every animator squash and stretch, anticipation, follow-through and overlapping action, exaggeration, and so on, as though these were animation’s laws of physics. Animation historian Maureen Furniss’s classic introduction to animation theory is called Art in Motion: Animation Aesthetics, not “Art in Stillness.” Movement is the central concern of animation as an artistic practice.
When we shift our interest in slow cinema from live-action images to animation, we thus make a discovery that is both surprising and entirely reasonable: an aesthetics of slowness is almost wholly absent not only from animation criticism but also from creative practice. Typical American animation follows the twelve principles and must maintain an intense pace and density of movement. It is therefore often animated on ones—one drawing per frame—and cannot fall below twos—one drawing per two frames. This is what is called full animation. American animation is a world in which animators such as Disney’s Nine Old Men become famous while directors retreat behind the scenes. Its basic criterion for “good” animation is a continuous flow of rich, varied movement, one action following another without pause.
Japanese animation, by contrast, developed a wholly different industrial and aesthetic system. The American industry consisted of large numbers of shorts and a small number of high-budget features. In postwar Japan, financial, staffing, and technical constraints meant that, from its very inception, the animation industry strictly controlled drawing counts to make production sustainable. It was willing to lower animation quality drastically so that work could be mass-produced for television. Japanese animation relinquished the aesthetics of movement from the start. Limits on drawing counts made it fundamentally impossible for Japanese animators to follow the twelve principles. They could only consider how to improve expressive power within a low-cost framework, developing a range of stylistically diverse aesthetics through limited animation. This ultimately produced today’s Japanese animation industry, with its rich variety of artistic practices and its intensely cult-oriented late-night anime market.
Japanese animation is therefore unavoidable in any discussion of animation’s aesthetics of slowness. Since the 1950s, the industry has produced a vast body of work characterized by stillness, slowness, and delay. Some of it has long been renowned in international art cinema: the many animated films of Mamoru Oshii, for example, whose work was nominated at Cannes in 2004. Other works became much-discussed commercial hits, such as Naohito Takahashi’s To Heart (1999) and Steel Angel Kurumi Zero (2001), and Shinya Kawatsura’s Non Non Biyori (2013) and Non Non Biyori Repeat (2015). Taking these three filmmakers as its points of reference, an aesthetics of slowness as its connecting thread, and mise-en-scène as its axis, this essay will analyze their varied experiments and expressions of authorship within slow anime. It will also indicate the aesthetic characteristics and medium specificity that distinguish slow anime from slow cinema.
I. Space—Realism
Animation has never seemed a naturally realist medium. While André Bazin sang the praises of cinema’s realist phenomenology and long-take aesthetics, animation was understood, at its core, as an art of fantasy. This distinction reaches back to the ontological roots of painting and photography. Perspective was an invention in painting from the beginning; realism was not painting’s nature but an artistic pursuit of painters. The invention of photography prompted an unprecedented debate about realism in the arts. Fantasy and abstraction eventually became the directions painting would take to distinguish itself from photography, as though the image-making machine had acquired a permanent monopoly on the duty to represent reality.
Animation is the art of fantasy. Ladislas Starewitch praised the animator who “roams tirelessly through the kingdom of fantasy. At will, he can reshape anything from the vast, chaotic heavens to a body a hundred-millionth the size of the tiniest insect.” For Loyd Price, “The beauty of animation lies in giving life to something that does not move. Cinema is an art made for fantasy, but it is in animation that fantasy reaches its peak. Once you take the camera away, after all, what you have in front of you is just a pile of plasticine.”4
Yet animation is by no means an image form that is ontologically incapable of realism. The contribution of time to the realism of the moving image is far clearer in animation than in live-action film. Because live-action images directly record reality, we more readily see their faithful reproduction of it and treat them as moving photographs. Animation remains essentially drawing. It lacks this close relationship between signifier and signified and may be called an incomplete image-sign. Yet precisely for this reason, the temporal character of the moving image becomes apparent. The most crucial difference between live action and animation, on the one hand, and photographs and paintings, on the other, is that moving images provide an entirely new dimension of time. A photograph offers a “real unreality” (irréalité réelle): we see a space presented as “here” (ici), but it is only a slice of “then” (autrefois). Moving images are different. Whether or not viewers recognize them as records of the past, the time they experience always unfolds before their eyes in the present tense. The reality of “it’s me” (c’est moi) overwhelms “this was so” (cela a été): “the having-been-there gives way before a being-there of the thing.”5
The real difficulty for animation’s realism lies not in time but in space. “The cinematographic image can be emptied of all reality save one—the reality of space.” Bazin adapted this assertion from Henri Gouhier’s formula for theater: “the stage welcomes every illusion except the illusion of presence.”6 Cinema does not, as theater does, structure the real presence of characters through a conventional space. The human being is far less central to cinema than to theater: “The human being is all-important in the theater. The drama on the screen can exist without actors.”7 In cinema, space is the true protagonist. “For a time, a film is the Universe, the world, or if you like, Nature.”8
In The Anime Machine: A Media Theory of Animation, Thomas Lamarre places the realization of space in live action and animation within a binary opposition between closed and open compositing. In cinematism’s closed compositing, space is integrated “seamlessly” into the standard Cartesian model. Changes from image to image are geometrically grounded in a consistent, uniform space. In animetism’s open compositing, by contrast, multiple planes disrupt spatial closure from within. Cel animation composites spatial depth from multiple layers on a flat surface. Changes in the relative positions of graphic elements create an illusion of depth, but actual movement into depth is rare.9
Spatial reality depends on closed compositing, and that is exceptionally difficult for animation. Like perspective in painting, spatial reality in animation is an invention through and through. This was the difficulty Mamoru Oshii faced when directing Patlabor 2: The Movie (1993). During the production of the two Patlabor films, Production I.G was still a newcomer, with few experienced staff members. Only Kazuya Kise made a deep impression on Oshii.10 The first film had a tight budget and schedule; the staff were run ragged, and Oshii himself did not consider the result satisfactory. He therefore made the second film his “revenge,” vowing to turn things around in one stroke. He virtually began learning how to make films from scratch, even asking where the 16:9 aspect ratio came from. Since I.G had no animation paper suitable for a 16:9 film, Oshii had to make it himself.
Through the discipline of setting aside all his previous experience and tracing cinema back to its essence, Oshii completed his methodology for making animated films. It encompassed formulas and lessons drawn from experience, ranging from film grammar to the practical work of drawing and animation, which he set down in Methods: Mamoru Oshii’s Patlabor 2 Direction Notes. He would make frequent use of this methodology in Ghost in the Shell (1995) and Avalon (2001). It became his internationally recognized signature and exerted a profound influence on the animation industry that followed.
The style of Patlabor 2 can be summed up in one phrase: “animation that does not move.” It earned this description through its abundance of static long takes and the scarcity of character movement throughout the film. Characters either stand still or move in regular patterns that do not disturb the composition. Why? Because Patlabor 2 is an animated film governed by composition. To achieve his “revenge,” Oshii prepared as much material as possible before production, devoting considerable thought and time to the layouts—the designs that determine each shot. Layout work took five months, while the actual animation was completed in just three. To keep the finished shots faithful to the layouts, neither camera nor characters could move enough to disrupt the composition.11 At a time when traditional animators still regarded movement as animation’s “life,” Oshii, intentionally or not, actually “castrated” movement, transforming animation into space. Whether he intended it or not, Oshii probably could not have imagined at the time how thoroughly this pioneering work would transform Japanese animation.
Oshii has studied and thought extensively about film theory and published a great deal. Research into his image aesthetics and mise-en-scène therefore often refers to his own methodology: the famous “three-distance method” from his Patlabor 2 Direction Notes, for instance, which links different shot scales to different narrative elements. Yet for that very reason, academic study of his images has seldom escaped the limits of his own account. If we look beyond animation to live-action cinema, it is not difficult to see the interesting parallels between Oshii’s style in this period and the Taiwanese New Wave. Long takes in full or long shot, sparing use of editing, characters who move little or not at all: these are the very aesthetic characteristics for which Taiwanese New Wave filmmakers are known, and they are also Oshii’s visual preferences. They resonate, too, with the broader field of New Asian Cinema.
In Patlabor 2, Oshii reduces the number of shots as far as possible. Where narrative grammar does not require a cut, he keeps the shot intact wherever possible. Early in the film, we see two scenes introducing different principal characters: Asuma Shinohara and Noa Izumi riding a bicycle (06:16–07:09), and Shinobu Nagumo taking an elevator after finishing a briefing (12:53–14:20). Apart from the transitions before and after each scene, both use just two shots, one long and one short. The short shot brings the viewer into or out of the long take, effectively serving as another transition. Strictly speaking, then, Oshii gives each scene only one principal shot. An eye-level tracking shot follows Asuma and Noa on the bicycle, the background flying past to create a vigorous visual rhythm. Shinobu, by contrast, is held motionless inside the elevator by an extreme high-angle shot, suggesting her dejection and inner stagnation.
An even more pertinent example is the briefing before the final battle (1:31:35–1:31:53). This is one of the few scenes in Patlabor 2 with many characters. Here too, Oshii chooses to use few shots, performing the narrative work through staging within the frame. At first, Ota’s face is completely hidden (fig. 1). When he begins speaking, he moves his head slightly to bring his eyes into view (fig. 2). Before Ota has finished, Shinshi, in the right foreground, turns around, revealing Ota’s entire face, then takes over speaking (fig. 3). Asuma raises his hand to ask a question (fig. 4). Izumi immediately turns her head toward him, blocking Yamazaki’s face in the process (fig. 5). Finally, when Captain Goto’s name comes up, we see every character’s reaction (fig. 6). Through complex staging of occlusion and a game of visibility, the composition is repeatedly remade within a single shot. Every change either focuses attention or draws a connection between characters, doing precisely what the staging requires.
Borrowing a term from psychology, David Bordwell calls this kind of mise-en-scène “just-noticeable differences,” or JNDs.12 The technique consists of full and long shots, fixed camera positions, and long takes. Traditional practitioners of the long take, such as Orson Welles, tend to establish a conspicuous visual center when staging in depth, placing the most important narrative element in the foreground. With just-noticeable differences, however, dense compositions and minute adjustments within the shot do the narrative work. They either redirect viewers’ attention without their noticing the staging’s precise guidance, or compel them to change how they watch: to observe actively rather than passively absorb what the film presents and be overwhelmed or moved by it, as in the classical mode. Yet Bordwell made this analysis not of Oshii, but of Hou Hsiao-hsien. In his book, he regards this technique as Hou’s greatest contribution to the formal language of cinema.13
There are many fascinating points of comparison between Oshii’s and Hou’s ways of composing images. While Hou was filming The Boys from Fengkuei (1983), his friend Chu Tien-wen lent him a book by Shen Congwen to encourage him. Unexpectedly, it gave him an entirely new way of seeing. In Shen’s writing, the grand narratives of history are always kept at the edges of the main narrative, while everyday life remains at its center. Shen “never preaches or passes judgment,” surveying the world from a distant, detached perspective.14 To capture the same feeling, Hou used long and extreme long shots that included the full figures and their surroundings for roughly forty percent of the shots in The Boys from Fengkuei. Comparing Hayao Miyazaki’s and Oshii’s directorial styles, Toshio Suzuki observes that Miyazaki “moves the audience through emotionally expressive protagonists.” Oshii works quite differently: he “creates by observing the characters objectively, carrying expressionless performance through to the end.”15 Where Hou films his “subaltern history” from a distance, Oshii’s animated films write politics and history from a thinker’s position. Oshii prefers tighter framings—mostly full and medium shots—with several characters together to stage just-noticeable differences. Yet his characters always keep their distance from the city, looking at it from afar and thinking about it. The city is pushed into the background or framed outside a window as a flat visual sign, independent of and opposed to the characters: figure 7, Twilight Q: Labyrinth Objects, File 538 (Meikyū Bukken File 538, 1987); figure 8, Patlabor 2; figure 9, Ghost in the Shell. In animation, Oshii has always been known for static long takes—and perhaps also for being “long-winded,” “preachy,” and “boring.” The minute performances within his shots and his expository dialogue both aim to alter viewers’ aesthetic distance and their way of apprehending the relation between image and sound.
Both Hou and Oshii are concerned with postcolonial and political subjectivity, but they approach them differently. In A City of Sadness (1989) and The Puppetmaster (1993), Hou reckons with Taiwan’s postwar and prewar history through a personal view of history, portraying “from below” people outside the center of historical events. In Oshii’s Patlabor 2, concern for ordinary people survives only in landscape “sketches.” Its narrative axis cuts directly into the mechanisms of politics. Without compromise or imagination, it points straight at the system’s failings, confronting the ontological questions of “just war” and “unjust peace,” the mechanisms of simulation in “fictional reality” and “real fiction,” and the muddy dilemma of postwar Japanese politics oscillating between left and right. Yet both Hou and Oshii ultimately use an aesthetics of “looking from afar” and “arrested movement” to portray the individual’s powerlessness before institutions, history, coercive power, and the collective—whether that individual is an ordinary person in history or a thinker who keeps his distance from the system. Each has arrived at a contemplative way of writing in images: strikingly similar and profoundly different, yet perhaps reaching the same destination by separate paths.
Looking across world film culture in the 1980s and 1990s, we find European art cinema gradually entering an impasse. At the same time, the wide influence of neorealism and the New Wave was reaching film cultures across Asia, taking root in mainland China, Taiwan, Japan, Iran, and elsewhere. With reform and opening up came the rise of mainland China’s Fifth Generation directors, who won award after award at major international festivals. Taiwan’s cultural and political climate also began opening up in the 1980s.16 Together with a crisis in the film industry and the stimulus of the Hong Kong New Wave,17 this helped bring New Cinema into being. The Japanese New Wave began in the 1950s and had largely run its course by the 1970s. In Japanese animation, however, the realist trend arose with Mobile Suit Gundam (1979). During the 1980s, Japanese cinema brought a new set of styles into animation. Yoshiyuki Tomino, famous for the Gundam series, had studied film at university and originally hoped to enter the film industry. Unable to find work amid fierce competition, he eventually joined Osamu Tezuka’s Mushi Production. Likewise, the cinephile Oshii set his sights on becoming a film director as soon as he graduated. He too could find no job and wound up at Tatsunoko Production almost by accident. Satoshi Kon, later renowned for his editing, was also absorbed in foreign films during his university years, an experience that deeply influenced his subsequent style.
Steeped in cinema, these animators displayed a wealth of invention in their work, bringing Japanese animation a new visual style distinct from Osamu Dezaki’s animated expressionism. In his “Layout History Course,” Shinsaku Sasaki divides this development into two periods, with the turning point around the beginning of the 1980s. Before the 1980s, the representation of space in animation gradually developed from an abstract, planar, symbolic space into a mature cinematic space organized in depth. From the 1980s onward, through the impact of works such as Royal Space Force: The Wings of Honnêamise (1987), Akira (1988), and the Patlabor series, varied ways of depicting urban space became established, while modern cinema’s handling of the camera and focal length was introduced and refined.18 The resources for this innovation came from the industry’s cinephiles. Oshii, who made his directorial methodology fully public in the Patlabor 2 Direction Notes, was a leading force in it.19
Oshii, a lover of European cinema deeply influenced by Ingmar Bergman and Andrei Tarkovsky, and Hou, whose ideas were challenged by Taiwanese filmmakers returning from abroad, such as Edward Yang, share creative similarities yet differ profoundly in substance. They are two representatives of the creative explosion in Taiwan’s film industry and Japan’s animation industry during the 1980s and 1990s, when Asian film cultures were following the example of the European New Waves. Where live-action New Waves broke with cinema’s classical mode, the rise of realism in Japanese animation marked a thorough departure from Dezaki-style expressionism. As the introduction argued, Japanese animation relinquished the aesthetics of movement from the outset. Dezaki’s system overcame straitened production circumstances through expressionist methods, using different directorial techniques to transform limited animation into a new kind of “dynamism.” The realist trend of the 1980s took a more radical step. Staging within a single shot replaced the extensive use of analytical and constructive editing in earlier animation. Bodies, actions, and space were no longer broken up by narration; space was established as the image’s true “protagonist.” I call this the “spatial turn” in Japanese animation. This turn brought animation closer to cinema’s artistic tradition, enabling Japanese animation to inherit live-action techniques and transform them into aesthetic effects peculiar to animation.
In The Sky Crawlers (2008), Oshii further developed his staging of just-noticeable differences. During production, he instructed Tetsuya Nishio: “Do the trick of ‘moving while remaining still’”; “Don’t make it completely still. The point is to create an impression of stillness.”20 What does that mean? Put simply, something looks completely motionless at first glance, yet contains extremely minute performance. Viewers feel that the image is still, but if shown an image that actually was still, they would feel quite differently. Oshii’s animated films often place characters in situations where they lack agency. They stand apart from the situation as seers rather than agents; the sensory-motor schema breaks down in the image. The minute performances in Patlabor 2 are often crucial to understanding the plot. In The Sky Crawlers, performance seeps into the viewer’s unconscious. This mode of expression takes the characters entirely outside the plot-and-character dominance of classical narration. “Atmosphere” turns the tables, becoming central to understanding Oshii’s films. Oshii imports cinematic techniques into animation, but transforms them completely. As everyone knows, he is revered as the other great figure in animation alongside Miyazaki; in live action, however, he has suffered repeated failures and produced few films of distinction. For a man who loves cinema, this is surely ironic. The reason is that, although his directorial techniques come from cinema, they can only be realized through animation.
Consider a conversation between two characters in The Next Generation: Patlabor—Tokyo War, Oshii’s 2015 sequel to Patlabor 2. The two characters are completely separated on opposite sides of the frame by the rule-of-thirds lines. Gotoda, the protagonist on the left, has his face in relative shadow (fig. 10). Takahata, on the right, then shifts a little to the left, while Gotoda’s attention wanders to the aquarium (fig. 11). In both figures 10 and 11, the aquarium sits at the lower-left intersection of the thirds lines so that it can become the center of interest as the scene develops. Gotoda’s face becomes more brightly lit as he lowers his head. Takahata then notices that Gotoda is not listening. The shot in figure 12 provides a transition between compositions and a pause in the conversation. The scene is then recomposed: Gotoda and the aquarium straddle the left thirds line, while Takahata moves onto the right one and dominates the image (fig. 13).
As what amounts to Oshii’s live-action remake and new-era reinterpretation of Patlabor 2, Tokyo War contains many passages whose events closely resemble those of the earlier film, and its filming techniques likewise approach Oshii’s early visual style. Yet it has received little praise since its release and has widely been regarded as another failure in his live-action practice. To a large extent, the problem is not a failure of mise-en-scène. The real issue is the essential difference between live action and animation. The changing light on Gotoda’s face is one of the keys to this scene’s staging, but its treatment in the finished film is very subtle. In animation, the same changes of shadow, composed of clearly bounded lines and planes, would be clear and forceful.
One of live action’s greatest advantages over animation is its mystery and inherent complexity. Take the same close-up of a face: a real face is always undergoing minute changes, and its forms are more complex. All told, it contains an enormous amount of information. Viewers can stare at it for a long time without necessarily becoming bored. Something inexpressible is always there. An animated face, by contrast, is drawn and stands closer to a concept than to reality. Discussing the relationship between live action and computer animation, Satoshi Kon remarked: “The spread of computer animation may simplify image-making in live-action films. If masterpieces such as Kurosawa’s Seven Samurai, Coppola’s Apocalypse Now, or David Lean’s Lawrence of Arabia had been made with computer animation, wouldn’t their images be far less powerful than the originals? Somehow, computer animation always seems to tend toward narrowing the film image.”21 Kon is referring here to computer-generated effects, not Japanese animation. Even so, human-made things seem inherently unable to attain the richness of the external world.
If filmmakers know how to use this, it can actually become an advantage for animation. Most animation cannot adopt an especially complex visual style if its images are to move. Unless the makers really have time and money to spare, high-budget Disney features, or works like The Old Man and the Sea (1999), which took two and a half years to produce just twenty minutes, are far too extravagant for either the animation industry or individual artists. We can therefore say that animation has little choice but to simplify. Its expressive power lies in lines and planes—in drawing. One way to realize the medium’s possibilities is to use lines and planes for precise visual distillation and tension in stillness. This is exactly how Oshii’s long takes work. They do not work like live-action long takes, which respect how their subjects naturally present themselves and give us an immediate, intuitive sense of reality. By maintaining the integrity of space and time, animated long takes allow delicate performance, precise layouts, distilled shapes and shadows, and the complex painterly textures of exquisite backgrounds to reveal themselves. Transfer the same techniques wholesale to live action, however, and the mismatch becomes plain. In absolute terms, live action cannot achieve animation’s precisely simplified light and shadow or its perfect control over the amount of information. Oshii’s just-noticeable-difference staging therefore ceases to work.
Tomomi Mochizuki once encountered this comment on Ocean Waves (1993): “If you’re going to draw it like this, why not just get real people to act it?” He replied: “If we made it as a live-action TV drama, we’d film a lot of extraneous things. The acting and the backgrounds, for example, would affect its texture. Animation has the advantage of being able to remove precisely those extraneous things.”22 Cinephile animation directors of Oshii’s generation often face questions like these: “How is your work different from cinema? Why don’t you go and make a film?” For Mochizuki, “That means what we’ve drawn has worked. Actually, it isn’t realistic at all.” “There are many things only animation can do. But then, those are also the things animation finds hardest to express.”23
Correction (added later): By “typical American animation,” I did not mean that full animation simply means animation on ones and twos. In my account, I still overlooked examples such as PUA and considered only Disney; I also concentrated only on television when discussing Japanese animation. This was unquestionably an error.24
Original editor’s note: Our thanks to Binggan for authorizing the republication of this fine essay. Two further sections follow.