Virtual Realism

In 2001 Chinese film critic and scholar Chen Xihe 陈犀禾 published an article in a leading Chinese film studies journal that celebrated the “postfilmic” world of digital cinematography.5 In that essay, which would soon be canonized by its addition to the two-volume anthology Selected Works of One Hundred Years of Chinese Film Theory, Chen praised computer imagery for providing filmmakers with a new “genuine freedom” in which cinema no longer rests on a foundation of ontological realism: “Digital imagery leads to the collapse of Bazin’s ontological theory of the image. As the digital image seizes the primary role in production, in fact any image at all becomes possible.”6 Thus “verisimilitude is no longer the goal,”7 and instead of photographic ontological realism, postfilmic cinema will be based on what Chen called the “virtual realism” of the artificial image. Similarly giddy pronouncements of the radically new world of digital media were made in the West around the turn of the millennium, particularly from theorists of new media rather than film, in what Philip Rosen calls “the discourse of the digital utopia.”8 As Lev Manovich put it in his seminal book The Language of New Media (published in the same year as Chen Xihe’s essay),

once live-action footage is digitized (or directly recorded in a digital format), it loses its privileged indexical relationship to prefilmic reality. The computer does not distinguish between an image obtained through a photographic lens, an image created in a paint program, or an image synthesized in a 3-D graphics package, since they are all made from the same material—pixels. And pixels, regardless of their origin, can be easily altered, substituted for one another, and so on. Live-action footage is thus reduced to just another graphic, no different than images created manually.9

Consequently, according to Manovich, “cinema can no longer be clearly distinguished from animation. It is no longer an indexical media technology but, rather, a subgenre of painting.”10

To quickly review what is at stake here, like many others, Manovich is drawing on the distinction first made by the pioneering semiotician Charles Sanders Pierce between symbolic, iconic, and indexical signs. In symbolic signs, such as human languages, the relationship between sign and referent is arbitrary; different languages thus use entirely different sounds to refer to the same thing. In iconic signs, such as a representational painting, the relationship between sign and referent is not arbitrary but rather one of physical resemblance; a portrait painting signifies its subject because it looks like that person. Finally, in indexical signs, the sign represents the referent through some sort of physical or causal relationship, as in the relationship of smoke to fire, a footprint to a recent presence, a thermometer to a certain air temperature, or a pointed (index) finger to the thing it points to. As the introduction and chapter 1 discussed, photography and film—which of course traditionally was made up of twenty-four still photos projected per second—have long been thought to have a particularly powerful effect of realism because the photographic image not only looks like its referent but is actually a direct imprint of light that has reflected off its referent at a past point in time; it is not only iconic but also indexical. What Manovich is arguing, then, is that with digital photography, the indexical nature of cinema is lost, and it becomes principally an iconic art again, like painting. The reduction of the image to digital code subject to infinite possibilities of manipulation by a computer compromises its cause–effect relationship with the external world and therefore makes even live-action photography potentially a mere tool of animation.

We can see the loss of photographic realism to which Manovich refers most clearly in films that make abundant use of CGI. Beginning in the early 1990s, not just Jurassic Park but also other films, such as Forrest Gump (Robert Zemeckis, 1994), were able to combine CGI with live-action cinematography in a way that made it difficult, if not impossible, to tell where one ended and the other began. This of course progressed to the point that a film like Avatar can begin with the so-called performance-capture of actors on spare, artificial sets and then manipulate the image so that the film as a whole appears as an incredibly elaborate and intricate cartoon fantasy world that nonetheless has a great deal of “real” photography seamlessly integrated into it.

Many examples also can be cited in Chinese cinema. We can turn to a scene from the blockbuster Red Cliff (Chi bi 赤壁; John Woo 吴宇森, 2008) to show that it is not just the objects in the image that might now be computer created but also some of the basic techniques of cinematography, such as camera movement. In one of the film’s most spectacular shots, the “camera” follows a released carrier pigeon flying from one military encampment to another across the Yangzi River, passing in the process a giant armada of ancient warships. If attempted with old-fashioned photographic film, this would be an aerial tracking shot of inconceivable technical difficulty, but of course it is nothing of the sort, as most of what we see in the image is computer generated. The multitude of ships is an illusion; there were only a few boats of different sizes actually built, and all the others were created digitally out of the visual information from those models.11 What seems to be camera movement might be more accurately described as image movement; algorithms applied to the digital scenery cause it to float to the edges of the screen in a manner that imitates a tracking camera, though no real aerial camera flew across the river as depicted. As realistic as the image looks, it is a computer-generated fantasy that took live-action photography as a starting point and manipulated it far beyond any event that ever appeared before the camera. Digital compositing is used to create the illusion of movement as captured by a virtual camera, with actual photographic images from a mechanical camera becoming mere elements to be agglomerated into the virtual image.12

In the context of this book, particularly the discussion of the long take in the preceding chapter, one conundrum raised by the pigeon-tracking shot in Red Cliff—as by other famous long takes in contemporary cinema, such as the seventeen-minute shot that begins Gravity (Alfonso Cuarón, 2013)—is whether such analytical categories as “camera movement” (or the more precise “mobile framing”) and “long take” (or just shot length in general) should be defined in terms of an event (or not) of actual shooting by a camera or the experience of viewing the shot as a spectator. If a camera movement is virtual, is it still a camera movement? A camera did not record its actual movement through space, but the spectator experiences a realistic sensation of movement. Similarly, a virtual long take does not require a camera really to shoot for a certain duration, but the viewer experiences the shot as continuous even if it is in fact composited.13

The answer to the question of the realism of the virtual camera movement or virtual long take depends on whether one’s measure is ontological realism from the time of shooting or perceptual realism as experienced while viewing. One theorist of visual effects, Stephen Prince, has argued strongly for the latter, insisting that film theory has put too much emphasis on film’s ontological realism and thus “construed realism solely as a matter of reference rather than as a matter of perception as well.”14 He argues that favoring perceptual realism over ontological realism (indexicality) better enables us to theoretically understand “some of the fundamental ways in which cinema works and is judged credible by viewers.”15 In his view, the overriding of “referential artifice” by perceptual realism is not new to digital filmmaking—because “the creative manipulation of photographic images is, of course, as old as the medium of photography”—but digital imaging nonetheless is revolutionary, because “it increases to an extraordinary degree a film-maker’s control over the informational cues that establish perceptual realism.”16 If a moving image gives perceptual cues that provide a convincing illusion of things existing in three-dimensional space, then neither a real camera nor an actual profilmic event is necessary to achieve an effect of realism.

This idea of viewing the moving image as constructed with the aim of perceptual realism rather than captured through the magic of photography’s ontological realism seems to have become widely accepted by filmmakers who now shoot digitally, even when they are not aiming at creating the imaginary worlds of fantasy or science fiction. For example, Ethan Coen has said that whereas shooting on traditional film forces the filmmaker to more or less commit to how a shot is going to look at the moment of shooting, shooting digitally is less like taking a picture than capturing information that subsequently will be turned into a picture: “when you’re capturing it on film, it’s actually in the grain of the negative, . . . and when you’re capturing it digitally, you’re just sort of recording pixels, all of which are negotiable later.”17

Digital Liberation and Chinese Tradition

The more constructed the digital film image becomes, the more its perceptual realism overlaps with what Chen Xihe celebrated as “virtual realism,” where “any image at all becomes possible.” Such a conception of cinema liberated from ontological realism compels us to revisit some of the basic questions of medium specificity and cultural adaptation raised at the beginning of this book’s historical journey. Chapter 1 showed that early Chinese filmmakers and critics felt an imperative to adopt an aesthetic of realism despite the anxiety that Chinese society and culture, with its traditional preference for expressionism rather than realism in art, was not well suited to a medium that seemed to rest upon modern, Western notions of mimesis and objectivity. We saw, for example, how the early Chinese film theorist Gu Kenfu emphasized that cinema’s realism demanded that actors be able to ride a horse, paddle a boat, drive a car, or even fly an airplane, in contrast to the actors in traditional Chinese opera, who had only to suppositionally signify such activities through pantomime, simple props, and stylized gestures and postures. Obviously, a shift to the virtual realism of the new CGI-enhanced story world means that actors do not need to ride, say, a Pandoran banshee in Avatar, since that action will be constructed by digital graphics animators after filming.18

Does the digital revolution in cinema raise the possibility of a rethinking of Chinese film aesthetics, in which matching the West (or “catching up with the development of world cinematic art,” as chapter 6 noted was an obsession for Chinese critical discourse on film in the immediate post-Mao period)19 is no longer a central concern? In one sense, in terms of the global competitiveness of the Chinese film industry, there is still a priority of developing the capacity to compete on equal technical and artistic footing with Hollywood blockbusters. As Yomi Braester has summarized,

The commercial mainstream regards the blockbuster as a standard toward which Chinese films should strive. Supporters of the blockbuster model have invoked a discourse of urgency, seeing a pressing need for overhauling the film industry to comport with Hollywood’s technological skills and achieving comparable levels of revenue. In the name of a viable domestic market, they established a strong link between digital special effects and box-office success.20

The drive to match Hollywood’s resources in creating CGI-filled blockbusters was reflected in the creation of the Animation School in the central government–funded Beijing Film Academy in 2000 and its subsequent rapid expansion, to the point that it became “one of the largest and most profitable” departments in the academy, with twenty-one faculty members and 358 students at the time of this writing.21 The training at the Animation School focuses in part on learning the most advanced techniques being used in Hollywood and elsewhere. When I toured the facilities in 2007, for example, one laboratory room was mostly given over to the making of models based on Tim Burton’s recent stop-motion animation feature Corpse Bride (2005), the animation and digital effects of which had been created by firms based in London (including MPC, mentioned earlier for its work on The Curse of the Golden Flower). It appeared that the students were learning exactly how it was done by painstaking imitation.

However, the investments in animation in China are not simply about being able to do top-quality imitations of cutting-edge techniques developed elsewhere, nor is the goal only to have Chinese animation workers or firms in demand as outsourced collaborators for popular Hollywood films, in the manner of some animation studios in India, the Philippines, and South Korea. The drive to make China a leader in the perceptual realism of digital animation and compositing aims to bolster the effort to increase China’s cultural “soft power” and even up what Michael Raine has called the long-standing “geopolitical incline” constituted by the global imbalance of film industry influence in favor of Hollywood, ultimately backed by U.S. political, economic, and military power.22 As the differences in animation aesthetics mentioned earlier show, the Chinese film industry seeks not simply to match Hollywood’s technological capabilities but to contest the artistic and cultural values it promotes by offering equally advanced cinematic expressions of Chinese aesthetics and values—reviving in the process Chinese cultural debates over Westernization versus the preservation of Chinese identity (or rather its modern construction through a selective mining of cultural history) that go back at least to the late Qing dynasty and never have a simple answer. Through an analysis of several key CGI-heavy Chinese martial arts films from the first decade of the 2000s, for example, Vivian Lee has theorized that they evince a new “digital imaginary” that achieves global appeal partly by transposing the exoticism of the Chinese cultural tradition into the international aesthetics of digital effects, thus transforming “the visual language of the martial arts film from a local cultural phenomenon to a transnational one.”23 The domestic and international successes of CGI-laden martial arts blockbusters like Crouching Tiger, Hidden Dragon (Wohu canglong; Ang Lee 李安, 2000), Hero, House of Flying Daggers, Kung Fu Hustle, and The Grandmaster (Yidai zongshi 一代宗师; Wong Kar-wai 王家卫, 2013) show that the Chinese martial arts digital imaginary became a significant part of the global cinema scene in the first two decades of the twenty-first century.

A more recent example of the successful (financially as well as artistically) deployment of digital imaging in the service of reenvisioning the Chinese narrative tradition is Monkey King: Hero Is Back—the 2015 adaptation of Journey to the West (Xiyouji 西游记), the fantastical Ming dynasty vernacular novel—which set a new earnings record for an animated film in China. Ying Zhu notes that the film’s competitiveness with Hollywood at the box office revives a sense of patriotic cultural self-assertion that harkens back to the 1920s, “when Chinese filmmakers cultivated kungfu films in pursuit of a ‘new heroism’ to compete with Hollywood for a slice of the domestic market,” and that the film’s “success has resurrected hope for the emergence of a unique Chinese style of animation” that is “rooted in the Chinese fine arts tradition.”24 Indeed, insofar as the new digital “virtual realism” extends beyond composited CGI elements in live-action cinema to pure animation, it renews a concern with developing a “national style” of animation that goes back to the mid-twentieth century.25

Here, however, I will focus not on animation films per se but on films that incorporate CGI and thus blur the boundary between live-action film (with its ontological and perceptual realism) and animation (with its perceptual but not ontological realism). According to Julie Turnock, films that feature a composited mise-en-scène, whether using CGI or predigital special effects, can be divided into two contrasting trends: a photorealist aesthetic, on one hand, and an obviously animated look, on the other.26 The former constructs images that emulate the look of live-action photography, giving the illusion that what is seen actually was photographed in the real world, whereas the latter creates an obviously fantastical, even cartoonish world that does not pretend to have been photographed in the traditional live-action sense. Some Chinese CGI blockbusters go more for the photorealist aesthetic—the virtual long take tracking shot flying over Cao Cao’s armada in Red Cliff being one example. On the whole, however, it seems that more Chinese filmmakers prefer the animation aesthetic rather than photorealist aesthetic in their use of CGI. Zhang Yimou’s visual style, for example, always has been readily described as “painterly,” and the CGI in films like Hero aims as well for expressionistic beauty more than the illusion of reality. Meanwhile, Stephen Chow’s effects added to live-action footage—in films from Shaolin Soccer (Shaolin zuqiu 少林足球; 2001) and Kung Fu Hustle to Journey to the West: Conquering the Demons (Xiyou: jiangmopian 西遊·降魔篇; 2013) and The Mermaid (Meirenyu 美人鱼; 2016)—intentionally look cartoonish, taking the animation aesthetic to a humorous extreme. Tsui Hark’s (Xu Ke 徐克) The Taking of Tiger Mountain (Zhiqu Weihu Shan 智取威虎山; 2014)—a coproduction of Hong Kong and mainland Chinese studios (as are Stephen Chow’s later-career films and many other contemporary Chinese blockbusters)—similarly embodies what Turnock identifies as the animation aesthetic, but in this case, with such contemporary digital tricks as obviously virtual camera movements, “impossible” camera positions, and changes in speed that freeze or extend dramatic moments through the kinds of “speed ramps” used in The Matrix (Lana and Lilly Wachowski, 1999) or 300 (Zack Snyder, 2006). The styles of these three directors—Zhang, Chow, and Tsui—are all quite different, but all three directors share a disregard for making the perceptual realism of their films’ CGI look like unembellished photographed images.

Prescriptive Realism Redux

The types of cinema that tend to rely most on CGI effects—such as contemporary martial arts films, action/adventure films, and science fiction films—also constitute current examples of prescriptive realism in Chinese cinema. In this book, prescriptive realism has referred mainly to the idealized visions of Chinese socialist realism, with its “combination of revolutionary realism with revolutionary romanticism” (chapters 4 and 5), but the last chapter also described a more recent trend of “capitalist realism” that is also in a prescriptive mode.

As Ying Zhu’s point about the “new heroism” cited earlier indicates, a key feature of prescriptive realism in its many different forms is the figure of the hero, who embodies and models ideal behavior. Chapter 5 traced how Communist heroes became so superhuman by the time of the revolutionary model opera and ballet films that they amounted to superheroes. Indeed, Hollywood superhero movies—in offering a stark moral universe, a clear grid for action, and even a provisional, if fantastical, metaphysics—are examples of contemporary prescriptive realism on a global scale. The contrast with the post–socialist realist mode examined in the last chapter is clear from Jia Zhangke’s 贾樟柯 statement that “I have never had any interest in making films about heroes or icons of any kind. This is something I feel quite strongly about. So you never see heroes in my films, and I have no interest in shooting any films about those idols most people look up to.”27 While largely absent from post–socialist neorealism, the Chinese cinematic hero survives and proliferates in the current forms of prescriptive realism.

The Chinese films of the reform era that most resemble the propaganda vehicles of the Mao era are the so-called main melody films, which themselves have increasingly taken the form of CGI-filled action movies. “Main melody” is the translation of the Chinese zhuxuanlü 主旋律, which could also be translated as “central theme” or “leitmotif.” These are films made on lavish budgets in government-owned studios with well-established stars and directors. Like their predecessors during the Mao era, they often focus on mythologizing key revolutionary moments in modern Chinese history. For example, The Founding of a Republic (Jianguo da ye 建国大业), released in 2009 to celebrate the sixtieth anniversary of the establishment of the People’s Republic of China, was a sprawling, star-studded epic directed by Huang Jianxin, who, in the 1980s, had been considered one of the leading figures of the Fifth Generation of filmmakers, known then for their subversive and experimental art films. The Founding of a Republic is on the contrary a hymn to the Communist victory in the civil war against the Nationalist Party (or Guomindang) and the birth of China as a Communist state. It took third place in the Chinese box office that year (just behind a Transformers movie), and its gigantic cast included a good portion of the top movie stars from both mainland China and Hong Kong (including some who are well known even in the West, such as Jackie Chan 成龙, Jet Li 李连杰, Tony Leung Ka-fai 梁家辉, and Zhang Ziyi 章子怡), although many appeared only in brief cameos. Many other Chinese “main melody” films similarly have been big-budget, highly patriotic narratives of key events in modern Chinese history, such as The Opium War (Yapian zhanzheng 鸦片战争; Xie Jin, 1997) and The Founding of an Army (Jianjun da ye 建军大业; Andrew Lau 刘伟强, 2017), which chronicles the origins of the People’s Liberation Army.

More recently, some of the most successful “main melody” films have focused on celebrating China’s growing military might in the Xi Jinping 习近平 era, when the country is increasingly flexing its muscles abroad, for example, by asserting its claim to the maritime territory of the South China Sea and investing in infrastructure throughout the Global South in what some critique as a new age of Chinese economic imperialism. The most notable recent examples have been Operation Mekong (Meigong He xingdong 湄公河行动; Dante Lam 林超贤, 2016), Wolf Warrior 2 (Zhan lang 战狼 2; Wu Jing 吴京, 2017), and Operation Red Sea (Hong Hai xingdong 红海行动; Dante Lam, 2018), featuring fictionalized Chinese military operations in Southeast Asia, sub-Saharan Africa, and the Arabian Peninsula, respectively.

Wolf Warrior 2 is notable because it combined its patriotic message with action melodrama so effectively that it became the highest-grossing Chinese film up to that time. Set in a fictional sub-Saharan African country, Wolf Warrior 2 depicts the powerful and apparently entirely benevolent Chinese presence in Africa, first through the characterization of its hero, a former elite special operations officer in the Chinese military named Leng Feng (played by action star and director Wu Jing), and second by the Chinese Navy, which comes to the rescue at Leng Feng’s urging when a group of mercenaries threaten to take over the country and kill countless innocent Africans as well as many expatriate Chinese in the process. As if to drive home the message of China as an effective force for good in a dangerous world (in contrast to the United States, whose military proves to be entirely ineffective as the fictional crisis worsens), several scenes feature Africans showing boundless gratitude to Leng Feng in particular and the Chinese in general. Indeed, the extremity of Leng Feng’s heroism recalls the romanticist exaggerations of the revolutionary-era films discussed in chapters 4 and 5 and helps to establish Wolf Warrior 2 and its kind as a successor to and extension of the sort of prescriptive realism found in the socialist realist classics of the Mao era.

Of particular interest is the way that Wolf Warrior 2 continues the deployment of melodrama in the service of prescriptive realism. The particular style of melodrama that gives the film its affective power is very similar to that of Hollywood’s Rambo film series that began with First Blood (Ted Kotcheff, 1982). In her essay “Melodrama Revised,” Linda Williams has shown how those “male action melodramas,” along with several other Vietnam-themed Hollywood films of the late 1970s and 1980s, served to purge American audiences of their guilt over the immoral Vietnam War and instead portray their American protagonists as innocent victims in melodramatic fashion.28 Like John Rambo in First Blood and its initial sequel, Rambo: First Blood II (George P. Cosmatos, 1985), Leng Feng in Wolf Warrior 2 has left the military, which had honed him into an almost invincible war machine, and tearfully nurses his grievances. Just as Rambo mourns the death of his friend and former comrade from cancer due to exposure to Agent Orange during the Vietnam War and more broadly laments that the war was ostensibly lost only because he and his brethren did not receive adequate support from the American government, Leng Feng likewise mourns the death of a dear comrade (killed in the first Wolf Warrior film) and goes to Africa partly to escape his painful memories of that loss and his own subsequent abuse by the corrupt head of a real estate company and the police who backed him. He works out his grief through masculine pursuits like drinking games and soccer matches on the beach in Africa. Like Rambo, he is portrayed “as a virtuous victim whose only motive is a melodramatic desire to return to the innocence of an unproblematic love of country and the simple demand for that love’s return.”29 The pathos of Leng Feng’s situation increases the audience’s strong identification with him when the film’s spectacular action sequences play out, particularly when he faces off directly with the film’s chief villain, an American mercenary leader called Big Daddy, an embodiment of evil and greed who kills indiscriminately. Near the film’s end, when Leng Feng finally prevails at the end of a seemingly interminable fight sequence, the geopolitical subtext comes to the fore when Big Daddy tells Leng Feng, “People like me will always be better than people like you,” and Leng Feng replies “That’s fucking history!” before proceeding to bash Big Daddy to death. Thus any potential moral reservation about China’s economic imperialism in Africa is swept away by the assurance that, after all, China is only finally reasserting itself as a great nation after nearly two centuries of humiliation and subjugation by Western imperialist powers, going back to the First Opium War of 1839–42. The implausibility of much of the action and dialogue of Wolf Warrior 2 in the end is obscured by the spectacle of its Hollywood-quality, CGI-enhanced action sequences; the emotional force of its melodramatic characterizations; and the familiarity of its prescriptive realist mode, which encourages identification with heroic ideals rather than recognition of ordinary reality.

Similarly prescriptive—and, in its contemporary form, prone to abundant use of CGI—is martial arts cinema, identified by Chris Berry and Mary Farquhar as belonging to the “operatic mode” of Chinese filmmaking, which they oppose to the sometimes more dominant “realist mode” more similar to foreign cinemas, such as Hollywood.30 Berry and Farquhar point out that the operatic mode goes back to the origins of Chinese cinema. In fact, opera films account for many of the landmarks of Chinese film history, including the first Chinese film, the first Chinese sound film, and the first Chinese color film. The rhythms of Beijing opera, including the alternation of spectacular acrobatic action with moments of stillness, also have long been characteristic of Chinese martial arts cinema. In short, anyone prone to cultural essentialism, in China or abroad, would have no problem identifying kung fu (gongfu 功夫) and martial arts (wuxia 武侠) cinemas as emblems of Chineseness, so that international blockbuster hits like Hero and Crouching Tiger, Hidden Dragon can be seen as raising China’s “soft power” via global film spectatorship. Around the same time those films were achieving global box-office success, many Western films began to incorporate elements of the Hong Kong martial arts style into Hollywood action cinema, particularly after the success of The Matrix, which employed veteran Hong Kong martial arts choreographer Yuen Woo-ping (Yuan Heping 袁和平) to design its fight sequences.

But why call martial arts cinema—or, for that matter, related contemporary genres, such as comic book superhero films—forms of prescriptive realism, thus putting them in the same category as socialist realist (or “revolutionary romanticist”) films? In fact, the genres share many characteristics. During the Cultural Revolution, one formula that was supposed to guide revolutionary narratives was the “three prominences,” which meant that the positive characters should stand out from the characters in general, a small group of heroic characters should stand out among the positive characters, and a singular hero should stand out as the most noteworthy model character of all. This formula applies equally well to most martial arts films: opposed to the irredeemably evil villain(s) are, first, the decent people in general, including ordinary people in need of help; second, a smaller group of highly skilled martial artists who either fight on the same side or, in some cases, fight honorably against each other if unavoidable circumstances require; and third, generally one supreme hero, a martial artist whose fighting prowess as well as moral character—displaying traits like courage, loyalty, and sympathy—is superior to all the other characters, reaching almost superhuman status. The same was precisely the case for the most heroic characters in the model opera films of the Cultural Revolution; recall, for example, Zhao Yonggang’s dispatching of twenty-five opponents in just a minute and a half of hand-to-hand combat in the revolutionary model opera film Fighting on the Plain, as described in chapter 5. Like the model opera films, the martial arts genre serves as a form of prescriptive realism that entertains mass audiences while reinforcing conventionalized and affectively charged collective values regarding morality and community belonging.

More genres lately also have exemplified the prescriptive realist label as used here. The Wandering Earth (Liulang diqiu 流浪地球; Guo Fan 郭帆, 2019) became the first mainland Chinese science fiction film to receive a great deal of international attention. Based on a story by China’s leading contemporary science fiction author, Liu Cixin 刘慈欣, it tells how a pending catastrophic explosion of the sun forces earthlings to come together to turn the whole planet into a spaceship in order to travel to another solar system and escape global incineration. The future Earth is imagined to be led by China just as effortlessly as Hollywood disaster films have long imagined Americans as habitual planetary saviors. The film contains clear echoes of previous forms of prescriptive realism. For example, it has a scene of grandiose self-sacrifice and martyrdom that seems to borrow from similar films from the height of Mao-era socialist realism, such as Dong Cunrui. In that film, a patriotic young soldier turns himself into a suicide bomber to protect his comrades from an enemy machine gun nest. Similarly, in The Wandering Earth, at a key moment when it looks like the entire planet will be destroyed, one of the main characters saves the day by turning the small spaceship he is flying into a missile, causing an explosion in Jupiter’s atmosphere that saves Earth but kills the pilot. In fact, this example resonates not only with Mao-era military films but also with the contemporary versions: the character who thus sacrifices himself is played by none other than Wu Jing, the director and star of Wolf Warrior 2 from two years earlier.

Aside from showing how science fiction can serve as a new vehicle for Chinese prescriptive realism, The Wandering Earth offers a case study in the differences in CGI aesthetics identified by Turnock, particularly when compared with the similarly themed 2014 international sci-fi hit from Hollywood Interstellar, directed by Christopher Nolan. CGI in the latter film follows the photorealist aesthetic favored in classic Hollywood sci-fi films like 2001: A Space Odyssey (Stanley Kubrick, 1968) and the original Star Wars trilogy (George Lukas, 1977, 1980, 1983).

One example of this aesthetic that is noticeable in Nolan’s Interstellar is the way that the CGI frequently incorporates fake lens flare (sometimes called camera glare) into the image. In conventional photography, lens flare occurs when a bright point of light strikes the camera lens, whether its source is in the image being photographed or out of the frame and hitting the lens from an angle, and the light is then scattered, or flared, within the lens system to create artifacts in the resulting photograph, from scattered glares that make part of the image look washed out to bright, semitransparent circles that occur across the image in straight lines. Obviously, no lens was involved in the creation of a computer-generated image, but such effects might be added in postproduction to reinforce the illusion of the image having been photographed in a real world. Berys Gaut identifies fake lens flare as one of the techniques of “photorealist” animation as opposed to just the “perceptual illusionism” of animation in general—a contrast that roughly coincides with Turnock’s distinction.31

Similar to this effect, and sometimes combined with it in Interstellar, is the very close attention paid to recreating the patterns of light and shadow when a continuously rotating object—such as the mother spaceship in the film—is hit from one side by a bright light, such as the sun. These moving images are meticulously designed to imitate the rapidly shifting patterns of shadow and bright light reflected off the surfaces of the illuminated spinning object. Such images occur repeatedly in the film.

Another aspect of Interstellar that illustrates the photorealist aesthetic is a kind of “imaginary GoPro effect,” referring to the miniature portable cameras that can be attached to moving objects, such as bicycle helmets, surfboards, cars, boats, or airplanes, to capture exciting action shots. Despite the fact that the camera is in fact virtual rather than real, the exterior space shots in Interstellar strive to give the illusion that the camera is somehow anchored to a physical object in the fictional world. In many of them, it appears that the camera is attached to the side of a moving spaceship. The point is to create the illusion that a camera captured the image from a physical position where it would in principle have been possible for a camera to be (Figure 30).

As China’s answer to Hollywood’s big-budget science fiction blockbusters, The Wandering Earth shares several plot points and motifs with Interstellar. Both are built on the premise of a dying planet Earth necessitating humanity’s relocation to another solar system; both share an obsession with fatherhood and with repairing an estrangement between father and child; both feature a spaceship with circular rotating parts; and both climax with breathtaking acts of self-sacrifice in which the main hero intentionally plummets to his apparent death—being sucked into a black hole in Interstellar and being consumed in a fiery crash into Jupiter in The Wandering Earth—to save others.32 Both, of course, also feature impressive CGI, but in The Wandering Earth, the aesthetic of the CGI shifts decisively from photorealism to animation.

In general, spectacular CGI-constructed exterior shots of spaceships, celestial bodies, and otherworldly planetary surfaces in The Wandering Earth mostly lack fake lens flare and, even when featuring similarly rotating spaceship designs, put less focus on realistically shifting patterns of light and shadow and more on the intricacies of the detailed illustrations of the fantastical images. Rather than constructing the illusion that its world was captured by a mechanical camera—eliciting a reaction of “It looks so real!” from the audience—the animation aesthetic aims more for a reaction of “Cool graphics!” of the type that one might have to a video game (Figure 31).

In particular, The Wandering Earth not only accepts but embraces the possibilities opened up by the presumption that, as Tobey Crockett has argued, “the mechanical camera is now effectively dead, and in its place the virtual camera rewrites both cinema and the world around us.”33 Its most spectacular shots do not seek to maintain the illusion that the “camera” exists physically in a real world but rather embrace the limitless possibilities of virtuality. As Roger Cook writes of such digital imaging, “there is rather a constant flux of unanchored positions, a dynamic flow in which space itself moves, bends, and morphs into various forms.”34

One shot begins with a close-up of the underside of a truck in motion (Figure 32a), but rather than appearing to be anchored to it in the imaginary GoPro mode, the virtual camera soars to the side and then up, within seconds turning the shot into an extreme long shot, and then keeps tracking impossibly back to reveal a vast mining operation on the surface of now-frozen Earth (Figure 32b). The virtual camera continues speeding back, soon showing one of the enormous thrusters now built into the planet’s surface to turn it into a giant spaceship, then follows the rocket exhaust beam up until the imaginary camera leaves what is left of the atmosphere (Figure 32c). Hardly finished, the “camera” continues tracking back through space until the entire Earth is a small sphere that is propelling itself away from the sun behind it, while the virtual camera flies along the side of the international space station that is escorting the planet to another solar system, until finally the focal perspective makes a cosmically scaled revolving track to the left until the space station, Earth, and the sun are three receding spheres arranged in a line in the vastness of space in the shot’s final seconds, when a title graphic also appears (Figure 32d). Thus concludes a virtual tracking shot that began as a close-up of the underside of a truck. The shot lasts fifty-five seconds in total, but within the first few seconds, it abandons any illusion that it has been photographed by a “real” camera accomplishing an ontologically “real” long take.

Another spectacular long take in The Wandering Earth is, during its first forty-seven seconds, a live-action virtuoso tracking shot filmed by an actual mechanical camera to show all the characters in the cab of the massive futuristic truck in which they are traveling (Figure 33a). However, for the rest of the one-minute shot, compositing allows the now-virtual camera to pass through the back of the vehicle—including its cargo, a “Lighter Core” (Figure 33b) to be carried to another of the planet’s “Earth Engines”—and exit the back of the vehicle (Figure 33c), which drives on into the distance, now in extreme long shot. Not only does the movement of the “camera” though multiple solid walls in the truck flagrantly display its own virtual status but the shot ends by adding a graphic display in the sky showing a map of the truck’s intended path from Hangzhou to Sulawesi, lessening further the shot’s photographic quality with an illustrated transmission of information to the viewer (Figure 33d).

Thus, in comparison to Interstellar, The Wandering Earth supports the thesis that CGI in Chinese cinema tends toward an obviously animated rather than photorealist aesthetic. In Turnock’s words, with an intentional twist on Bazin, such filmmakers put their “faith in animation” rather than putting their “faith in the optical.”35 In the animation style of CGI, “effects objects are presented from camera angles with no possible human perspective, and ‘cameras’ move through impossible spaces,” as we have just seen in the two virtual long takes in The Wandering Earth.36 Other effects in the film include manipulations of time, such as speeding up, slowing down, or freezing actions, sometimes combined with cameras that move while the action itself is slowed or frozen—the sort of speed-ramp effects mentioned earlier in the case of The Taking of Tiger Mountain. These are especially evident in a very late sequence in which characters plunge to the ground amid flying debris due to a shock wave hitting Earth.

Any discrepancies in CGI aesthetics between Chinese and Hollywood blockbusters may soon become blurred not only by the ability of Chinese filmmakers to imitate Hollywood but by Hollywood’s own efforts to court the Chinese audience. As the Chinese box office grows into the world’s largest—soon dwarfing that of the United States—the rest of the world is likely to become more familiar with China’s new forms of prescriptive realism, the aesthetics of which both borrow from and increasingly may be imitated by Hollywood itself, as the latter seeks increasingly to meet market demand in China to maintain its global status and profitability while coping with the sensitivities of Sino-U.S. political tensions.37

The Ontological Realism of the Digital Image

We began our discussion of CGI and its related forms of prescriptive cinema in China with the claims many have made that the change from photochemical film to digital cinema represents a fundamental break from cinema’s ontological realism in favor of perceptual or virtual realism. Are Chen Xihe and Lev Manovich correct in saying that, with digital photography, cinema loses its ontological tie to the real world? Many film theorists, such as D. N. Rodowick in his book The Virtual Life of Film, agree that the so-called indexicality of the image as been compromised, if not completely lost, with the rise of digital cinema. As Rodowick puts it, “in digital capture, the indexical link to physical reality is weakened, because light must be converted into an abstract symbolic structure independent of and discontinuous with physical space and time.”38 This conversion of visual (and audio) information into chains of digital ones and zeroes contrasts with reality inscribing itself directly “in the grain of the negative” (in the words of Ethan Coen cited earlier). Others, however, have questioned this logic, arguing that the switch from analog to digital photography does not necessarily mean a loss of the indexical link between image and world.

Tom Gunning, for instance, decries digital media’s alleged loss of indexicality as “nonsense” that falsely equates the indexical with the analog.39 He points out, for example, that medical devices have long converted information detected directly—which is to say, indexically—from the material world into numbers (such as a reading of someone’s pulse rate, temperature, or heart rate), and such a conversion does not suddenly destroy our faith in their reliability as measures of real physical states.40 Scott Curtis similarly observes that the difference in medical imaging between, say, an analog X-ray film image and digital images from a CAT or MRI scan does not lessen the reliability of the latter: “Even though the information gathered by the machines travels to the computer in the form of binary oppositions, ones and zeros, that information is nonetheless ‘indexical’ in the sense that there is a necessary physical connection—even if only at the molecular level—between the object and its representation. It must be so; otherwise, the images would have no informational value.”41 Philip Rosen likewise notes that contemporary military satellites used for digital surveillance generate data that are very much taken as recording material realities on the planet: “without referential entities or events preexisting the data itself, that data would have no informational value as surveillance.”42 Indeed, the modern Chinese domestic surveillance state—quite possibly the world’s most sophisticated—would be impossible if digital data were incapable of ontological realism, accurately recording things in the real world.

It takes only a little consideration of how digital imaging works in our daily lives to realize that the possibility of an image that looks photographic but is entirely computer generated—such as a computer-animated film in a photorealist style—hardly means that we should take all digital images as having no existential tie with the real world. When we take mobile phone snapshots, or even launch mass social protests based on a violent injustice that was caught on video by a bystander, we are taking as a given that at least some digital images are accurate records of something in reality, just as was the case with the analog photography used in the past. Berys Gaut concludes that while it is true that “digital imaging software makes available painting techniques that break the essential causal link between a photograph and its subject” as argued by Manovich, “the mere possibility of using such techniques does not show that, if they are not employed, the resulting image is like a painting, any more than the possibility of overpainting a traditional photograph shows that traditional photographs are like paintings.”43

All these examples show that what Rosen calls “digital indexicality” is not the oxymoron that so many film and media theorists implied in their turn-of-the-century anxiety (or jubilation) over the replacement of analog photographic images with digital ones and the ostensible ontological rupture that entails.44 Whether in the case of a military analyst inspecting a digital satellite image, a police detective viewing security camera footage, or the rest of us looking at photographs from our friends’ holiday trips on social media, the faith we put in the image’s connection to reality is based not specifically on whether it was generated photochemically or digitally but rather more generally on our assumptions about how and why the image was produced. Just as we expect to see computer-generated visual effects when we watch a special-effects blockbuster like The Wandering Earth, when we watch the latest gritty neorealist fiction film or cinema verité–style documentary from China, we expect to see some true vision of Chinese reality today, even if the footage was shot digitally.

Although documentary cinema lies outside the scope of this book, it is worth noting that it is precisely the digital revolution that has enabled the explosion of independent documentary filmmaking in China during the early decades of the twenty-first century. In his book on Chinese independent documentary, Luke Robinson notes that upon its arrival, digital video “was seen as unparalleled in its capacity to capture experiential reality,” amounting not to a loss of ontological realism but, on the contrary, to a “hyper-indexicality.”45 Similarly, in the introduction to their coedited anthology on the new Chinese documentary film movement, Chris Berry and Lisa Rofel argue that when digital video began to proliferate in China around the turn of the century, the discourse around the “DV aesthetic” emphasized its “on-the-spot realism” (jishizhuyi 纪实主义), so that the “Chinese understanding of the essence of DV stands in stark contrast to the common understanding in the United States and elsewhere in the West” as influenced by theorists like Manovich.46 This difference, they continue, “should alert us to the fact that DV has no single essence,” in contrast to Manovich’s medium essentialism that divides photochemical film from digital video or cinema.47 Gunning argues more broadly that we should view cinema in general “as a braid made of various aspects rather than a unified essence with firm boundaries” because it always has been “a point of intersection, a braiding together of diverse strands,” overlapping with many other technologies, from the phonograph in the early sound era to the computer in the current age of media convergence.48

In this light, the “digital indexicality” allowed by DV cameras—the opposite pole on the digital continuum from CGI’s “virtual realism”—has intertwined with the rise not only of the new documentary movement in China but more broadly of the independent film movement that, as the last chapter detailed, became the dominant form of post–socialist realism beginning in the 1990s. China’s independent filmmakers adapted quickly to the possibilities of new digital formats. Jia Zhangke, in the process of becoming the most internationally renowned director of the “Sixth Generation,” switched from Betacam video to 16mm film to 35mm film to DV just from his student films to his third feature film. He would not make a feature 35mm film after Platform (Zhantai 站台; 2000) until 2018’s Ash Is Purest White (Jianghu ernü 江湖儿女), so he has worked in digital formats for most of his career, even as he became known as a preeminent realist filmmaker. At the same time that he was entering his DV stage at the beginning of the 2000s, countless other filmmakers and, more accurately, videographers throughout China were answering a call-to-arms Jia had made in an essay published in the newspaper Southern Weekly (Nanfang zhoumo 南方周末) in 1999, in which he had called for the rise of an “amateur cinema” to contest the domination of industry professionals and official state film studios. Here Jia was not only referring to the genuine practice of amateur hobbyists but also using “amateur” more metaphorically to describe the new independent cinema of the 1990s of which he was a part. In fact, while working in the new “underground” sector of filmmaking, Jia as well as several of his fellow independent directors, such as Lou Ye, Zhang Yuan, Wang Xiaoshuai 王小帅, and He Jianjun 何建军, all were professionals who had been trained in the Beijing Film Academy. Over the following decade, however, Jia’s vision of the rise of “amateur” filmmaking would occur in a quite literal sense owing to the democratizing effects of the DV revolution. The resulting explosion of independent cinema has been far too rich, varied, and complex than it would be possible to detail here. Already mentioned was the prolific blossoming of independent documentary cinema. Alongside that, as was examined in part in the preceding chapter, in the 1990s and the first decade of the new century, many feature films were made in a gritty neorealist style that appropriated the same type of verité on-location videography as many of the documentaries. Corresponding to the independent film and DV craze was the rise of local film clubs and (for a time at least, although the authorities eventually cracked down on them) independent film festivals from Beijing to Shanxi to Yunnan.

Given the close link between the new documentary movement and the independent fiction film scene (Jia Zhangke himself has made a handful of documentaries, for example), it was no surprise to see an expansion of extremely low-budget feature filmmaking that blurs the boundaries between documentary and fiction. DV filmmakers would often forgo obvious postproduction possibilities to preserve mistakes or poor quality as a badge of authenticity, leaving visible such things as autofocus mistakes or temporary over- or underexposure problems when a DV camera automatically adjusts as it moves between bright and dimly lit spaces. For filmmakers shooting with nonprofessional actors, DV arguably allows for more realism because the director can shoot continually and potentially capture a larger number of “authentic” moments without having to worry about the imposing bulk of 35mm cameras, the expense of film stock, or the distractions of changing out film cartridges while shooting. 

Here I will briefly cite just a few examples of this phenomenon. The film Good Cats (Hao mao 好猫), directed by Ying Liang 应亮 in 2008, is a fictional account of local corruption and greed. It has an almost journalistic approach and does not disguise the markers of its amateur aesthetic, including occasionally poor acting and noticeably amateurish lighting and autofocus in the DV photography. While hardly a major artistic achievement, the film’s unflinching look at local greed and corruption earned it attention first in the independent film scene in China and then among international festival programmers and distributers. Other examples include films by the independent filmmakers Liu Jiayin 刘伽茵 and Song Fang 宋方, cited in the preceding chapter, both of whom made fiction films that seemed like documentaries because they were shot at the filmmakers’ family homes, with their own family members playing themselves in scenarios that seemed improvised, even though they were in fact substantially scripted. These two directors also exemplify a crucial consequence of the leveling effect that came with the drastic lowering of barriers to entry into the filmmaking profession brought by the relative affordability of the means of cinematic production: the DV revolution has done much to bypass the traditional sexism of the film industry (in China as elsewhere) and allow the cluster of talented young directors receiving attention in the domestic and international independent and art film scenes to include a higher proportion of women than at any previous time in Chinese film history.

Not only is the idea of a rupture in the essence of the medium with the coming of digital cinema overly simplistic, but even the more careful division I made between the CGI’s virtual perceptual realism and DV’s ontological realism implies a fissure that might easily be bridged in a particular film. The typical big-budget commercial film today, whether in Hollywood, China, or elsewhere, is likely to feature live-action footage composited—whether subtly and unnoticeably or grandly and spectacularly—with digital effects created on computer. At the same time, the worlds of the imagination conjured by virtual environments are also real parts of ordinary people’s lives, so that a film in a verisimilar critical realist mode might depict them as part of a quotidian reality it wishes to capture.

Hypermediation, Animation, and Digital Effects in the Films of Jia Zhangke

Jia Zhangke’s films are exemplary in this regard. From his first feature, Xiao Wu, despite the almost documentary style of on-the-spot realism he used to depict the lives of disadvantaged losers in a nondescript contemporary city in the Chinese interior, he and his sound designer, Zhang Yang 张阳, with whom he has continued to work throughout his career, were careful to depict various sorts of media—local news reports, overheard snatches of soundtracks from Hong Kong new wave films of the 1980s, off-tune karaoke singing blasting from distant speakers, and so on—as ubiquitous in the urban landscape, or, more important, soundscape. This kind of hypermediacy—the depiction of one medium within another—would become a hallmark of Jia’s particular brand of post–socialist realism.49 His second film, Platform, which traces a local performance troupe from the late 1970s to the end of the 1980s, realistically showed the degree to which popular music was incorporated into young people’s changing identities, including how pop music from far away (a love song broadcast from Taiwan via shortwave radio, a cassette tape of rock music brought back to Shanxi from the much more rapidly developing coastal province of Guangdong) could take on a utopian function in the imaginations of the youths. His third film and first shot on DV, Unknown Pleasures (Ren xiaoyao 任逍遥; 2002), featured several shots of a well-known cartoon adaptation of the Ming dynasty adventure novel Journey to the West, which one of the teenage protagonists of Unknown Pleasures habitually watches. This fantasy world provides a contrast to the bleak setting of the Shanxi mining city Datong, where Unknown Pleasures is set. While the characters’ real lives are severely constrained by limited opportunities, it is suggested, they can still yearn for a realm of freedom symbolized by the wild adventures of the Monkey King, liberated from reality by animation.

Such juxtapositions amount to the depiction of prescriptive realism within verisimilar social realism, so that the latter’s critique includes as its object not only the material lives of ordinary people under the conditions of postsocialist development but also the forms of media they consume that offer a romantic, even utopian vision of escape from those very conditions. As Jia himself put it,

the younger generation are faced with a new kind of cultural oppression. This is in part due to the lifestyles they hear and learn about through the media—especially the internet and cable television—which exists on a completely different plane from their everyday reality. It is this radical contrast between the reality of their environment and the picture of the world they get through the media that creates an enormous pressure in their lives.50

This contrast comes through even more starkly in Jia’s next film, The World (Shijie 世界; 2004), in which animation reappears, but in a way that is much more ambiguous in its relation to the world of the film in general. While the animation in Unknown Pleasures was clearly contained within the TV screen of a character’s apartment, that in The World enters in the form of cuts to animated fantasy sequences that take up the entire film screen rather than a diegetic screen depicted within the screen. All five of the animation sequences in the film are motivated by digital technology, specifically text messages received by the characters on their mobile phones. The Flash animations show hands holding the phones with the text messages on them, as if from the point of view of the real characters, but they also feature more oneiric images that seem to be flights of fancy or dreams of freedom or escape. In the first, for example, the main protagonist, Tao, is passing by Tiananmen on a public bus when we hear the tinkle of her ringtone. She lifts her flip phone to look at the screen, after which the whole screen switches to an animated virtual tracking shot of the bus from the front, then to a close-up of her phone (Figure 34) showing a text from her boyfriend (with whom she has recently fought) saying, “Where are you going to run to?” The scene then cuts to another virtual tracking shot, this time of the passing streetlights as if from the perspective of the bus itself. The sense of movement combines with the text message to convey both the attraction and the futility of finding a line of flight out of Tao’s fate as a migrant from a distant province working as an entertainer at a theme park in Beijing. The phones themselves, with their ability to collapse space and connect people across a distance, take on a romantic, idealizing function. As Sumanth Gopinath observes, “linked to the mobility of the cell phone and moments of transit that ultimately lead nowhere, Tao’s ringtone is arguably a little sonic lie about freedom in the face of the deadening routine that the park’s performers must endure in order to produce its dream-factory spectacles; that lie, however, is inseparable from the utopian horizon of an outside (and better, less alienated) world which the mobile phone seems to offer.”51 Later animated sequences in the film combine the cell phone interactions with, for example, a vision of Tao floating high in the air over the Beijing cityscape. Jia has related these fantasy-style representations of digital interactivity to the emergence of a diversifying internet culture in China at the time: “Suddenly young people found themselves living amid two worlds—the virtual world and the real world.”52 Just as the “world” theme park presented an ersatz fantasy world for its visitors, the internet offers an alternative virtual space that nonetheless becomes a real part of young people’s lives.

The undercutting of realism by the animation sequences in The World was a little shocking to fans of Jia Zhangke insofar as he had been taken to represent a relatively pure form of the neorealist aesthetic. However, as the preceding chapter detailed, by the early 2000s, that aesthetic was becoming conventionalized to the point of losing its power to provide a jolt of the real, instead seeming itself formulaic, and the use of animation to help construct the world of The World indicated a shift in Jia Zhangke’s style. While still concerned with revealing to the audience real social conditions elided in more commercialized films, he began to reflexively complicate his own apparent claims to realism, the first major example being the shifting of representation from the ontological realism of live-action film to animation in those five sequences in The World. The problematizing of his own style of realism continued in several ways in his following works from 2006, the fiction film Still Life (Sanxia haoren 三峡好人) and the documentary Dong 东, filmed at the same time. In Still Life, rather than animation sequences, Jia inserts isolated instances of obvious CGI in a film that otherwise appears to follow the usual markers of Jia’s neorealist style—on-location shooting, use of nonprofessional actors, long shots and long takes, and so on. In one shot that appears photographically realist in every other way, an otherworldly UFO suddenly streaks across the sky. In another shot, probably the most striking in the entire film, what looks like a half-constructed building in the background—actually a tower in the shape of 華 (hua, or China) built by a local government—suddenly ignites and takes off like a rocket ship headed for outer space (Figure 35). (As Jia Zhangke’s first foray into obvious CGI effects and second film to gain official distribution within China rather than being independent, Still Life nonetheless was sidelined at the Chinese box office by two of the CGI-filled blockbusters mentioned at the beginning of this chapter: Chen Kaige’s The Promise and Zhang Yimou’s Curse of the Golden Flower.)53

What are we to make of Still Life’s unexpected CGI inserts in terms of the practice of cinematic realism? First, the CGI effects obviously compromise the ontological realism of the photographic image in the manner suggested by Manovich. We know that there was not an actual profilmic event of the tower taking off in front of the camera but that instead the effect was achieved in postproduction through computer manipulation of data from the original photographic image of the building. Second, the perceptual realism of the tower’s movement nonetheless demands our “participation” or “immediate involvement in the image”; although the building in the real world did not move, in terms of the phenomenology of the viewer’s perception, there is a “real presence of motion.”54 Third, the verisimilar social realism of the film is of course compromised by the implausible visual effect. Buildings don’t suddenly take off like rocket ships, so such a depiction lessens the film’s otherwise strong sense of coinciding with contemporary reality—in this case, that of China in the settlements along the Yangzi River that were being demolished and flooded during the construction of the Three Gorges Dam project. Finally, the fictional realism of the film also appears to be severely challenged by this image, insofar as it potentially undermines the “worldhood” of the film as a whole. Even if we can attribute an allegorical meaning to the shot (China’s economy “taking off” or, alternatively, China losing its “grounding,” and so on), the fantastical nature of the CGI effects seems so out of place in comparison with the rather melancholy drudgery of the film as a whole that it almost appears to come out of a different movie altogether, rupturing the internal consistency of the film’s fictional world. At the least, the moments of sudden CGI surrealism would provoke the viewer’s engagement to switch from fiction emotions to artifact emotions (see chapter 5), shifting attention at least partly from the film as a fictional world to a constructed artifact,55 insofar as the first-time viewer’s immediate reaction was likely to be “What is an obvious CGI effect doing in a neorealist Jia Zhangke film?”

Jia himself counters that surrealism is a form of realism if the reality on which it focuses is itself surreal. In an interview cited and translated by Tonglin Lu, he insists that “a surrealist atmosphere prevails in China today” under the condition of modernization at breakneck speed, so that “the surrealistic elements . . . are part of reality.”56 In a later interview, he elaborated on the surrealist moments in Still Life, including the unexplained presence of a tightrope walker in one scene and the appearance of Sichuanese opera actors in full costume and makeup playing handheld video games in another scene, in addition to the CGI flying saucer and the building launching like a rocket ship:

My rationale for placing these surrealistic elements into a film that was otherwise rooted in realism was because the dramatic pace of China’s development during that time often felt as if it was somehow not real—it all felt surreal. Actually, what I most wanted to capture at the time was this surreal feeling that people living in that environment must have been experiencing. . . . Just because we were telling a story about the real world in contemporary China doesn’t mean we can’t employ surrealistic methods. As long as you are free, you can employ whatever techniques you like. And even when you do utilize these methods, lurking behind them there is still a clear realist intention.57

The scenes of massive demolition required for the construction of the Three Gorges Project, when entire buildings disappeared in an instant, made the idea of the building suddenly taking off into the sky seem only slightly more surreal than what Jia was in fact seeing around him on location every day during the shooting of the film.

Another way in which Jia complicated his previous realism was through the mixing of fiction and documentary in both Still Life and the related documentary Dong. In fact, neither film is either fully fictional or fully documentary. Still Life includes actual scenes of demolition in the Three Gorges area within its fictional world, with real demolition workers serving as extras in the film, for example. The documentary Dong, on the other hand, ostensibly a record of the painter Liu Xiaodong composing massive oil paintings from life in the Three Gorges area as well as in Thailand, has footage of a demolition worker who, as we would only know from watching Still Life, is in fact a semiprofessional actor (actually Jia’s cousin, who had acted before in Platform) who portrays a character in the fiction film. Consequently, the documentary Dong includes elements of fiction, while the fiction film Still Life contains documentary footage.58 Jia’s next feature film after those, 24 City (Ershisi Cheng ji 二十四城记), an apparent documentary which in fact seamlessly mixes scripted interviews with actors along with unscripted interviews with real subjects, carried Jia’s problematization of realist representation to a new level with its free mixing of fiction and documentary.

One thing Jia seemed to be doing—no doubt in part in reaction to the way the neorealist aesthetic of his early films had become so widely imitated and thus conventionalized—was to question his own role as a realist artist. Whereas his early films seemed suffused with the confidence that a filmmaker with sufficient courage and vision could go out in the streets of China and document in fiction films those realities that seemed to be omitted from mass media representations in general, with these films, Jia seems to interrogate his own subject position as a relayer of such images. Just as the documentary Dong in part calls into question the relationship between the elite painter Liu Xiaodong and the proletarian subjects of his paintings, Jia himself seems to acknowledge that his status as an internationally renowned film auteur inevitably puts him at a remove from the ordinary, underprivileged workers whom his films are generally about. The contradictions of ontological, perceptual, social, fictional, and prescriptive realist representation presented in The World, Still Life, Dong, and 24 City thus serve as acknowledgments that an art filmmaker representing the “real” China on the global scene cannot be a simple or transparent process.

Mixing Mediums, Genres, and Styles

As this and the preceding chapter have made clear, the mixing of fiction and documentary has been a consistent tendency of Chinese independent and art cinema, beginning with Zhang Yuan’s Mama (妈妈) in 1990 and continuing with a number of the most inventive filmmakers of the DV generation. Cutaways to actual documentary footage within fiction films go back to the classics of 1930s Shanghai cinema, as we saw in the case of Street Angel in chapter 2. In recent years, even Stephen Chow’s comedy blockbuster The Mermaid included documentary images of dolphins being slaughtered, which were combined in somewhat shocking fashion with the fantastical CGI that filled much of the film. Rather than seeing the former as ontological realism and the latter as formalist animation with no connection to reality, we might broaden the idea of the indexical and consider the film as a whole, including its CGI animation, as an index of the rising anxiety of those facing the ecological disasters caused by capitalist development and climate change.

Many other recent films have also mixed mediums, genres, and styles in inventive ways that call into question easy dichotomies, including documentary versus fiction, realism versus fantasy, photography versus animation, and even modern versus traditional. Emperor Visits the Hell (Tanghuang you difu 唐皇游地府; Li Luo 李珞, 2013) depicts the famous episode from chapters 9–11 of Journey to the West, but instead of animation or the usual elaborate costumes, prosthetics, props, and special effects of the countless previous film and television adaptations of the fantastical Ming dynasty novel, it is instead reimagined as a deadpan low-budget satire set in contemporary times. In keeping with the pseudo-documentary neorealist aesthetic of earlier Chinese independent cinema, the roles were all played by “friends and friends of friends” of the director, including punk rock musicians, artists, and so on—the lack of professional actors being one thing that kept the cost “very low.”59 Shot off and on over little more than a month, the film’s effect of having figures like a yaksha, the Dragon King, and the King of Hell played by ordinary people in China today is humorous, and the contrast between the fantastical events and the completely prosaic settings, costumes, and delivery is what gives the film its charm and originality. The final sequence switches into a casual home-video mode and has the actor who plays the main role out of character and drunk at a meal, ranting about avant-garde artists, university professors, and vaguely identified effeminate poseurs, not to mention putting down China as a whole. The rather strange experiment shows how traditional Chinese fantastic tales can lend themselves not only to digital wizardry but to the most stripped-down kind of DV “on-the-spot” realist aesthetic.

Several other contemporaneous Chinese independent films combine neorealist techniques with surreal fantasy. My Dear Friend (Haoyou 好友; Yang Pingdao 杨平道, 2018) mixes rural drudgery with magical realism. Bi Gan’s 毕赣 Long Day’s Journey into Night (Diqiu zuihou de yewan 地球最后的夜晚; 2017) complicates its indie neorealist aesthetic with a spectacular and surreal 3-D long take of nearly an hour that concludes the film. The Oxhide films of Liu Jiayin, as described in the preceding chapter, merge an extreme neorealism (on-location shooting, nonprofessional actors, almost plotless ordinary life) with an extremely formalistic stylistic patterning (moving the camera position by forty-five-degree increments in a rigid pattern).

Another radical formal experiment forged out of the most ordinary, even ugly, video footage is Dragonfly Eyes (Qingting zhi yan 蜻蜓之眼; 2017), by the globally renowned visual artist Xu Bing 徐冰, who assembles a mystery story told entirely through actual found footage from video surveillance cameras, with dubbed dialogue. Security or surveillance videos provide a kind of limit case of the idea of digital indexicality. As Ling Zhang argues in an insightful essay on Dragonfly Eyes, digital surveillance camera footage is an ultimate case of what André Bazin celebrated as photography’s absence of human agency in creating an image: “an image of the outside world takes shape automatically, without creative human intervention, following a strict determinism.”60 As Ling Zhang elaborates, “on the ontological level . . . footage generated by surveillance cameras might be regarded as totally self-generating and autonomous, a kind of ultra-realism, since it is devoid of any human intervention” beyond the original installation of the camera.61 Given the utter lack of interest or artistic intent in most surveillance footage—which, Zhang points out, paradoxically is made not to be viewed, as long as nothing abnormal occurs—Xu Bing’s transformation of it into narrative and film art is especially audacious, creating the first ever feature fiction film consisting entirely of such footage.

As is evident even from the limited set of examples described in the last few pages, the transition from celluloid film to digital video, rather than constituting any sort of simple dualistic rupture, instead involves a diffusion of new possibilities that do not so much negate cinema’s ostensible ontological realism as incorporate it into the threaded possibilities that constitute cinema in the digital age. The “indexicality” of cinema is not just a simple matter of whether a camera “automatically” records a profilmic event but rather asks a far more complex question of whether a film indexes an engagement with the actual, interconnected conditions of sociality, mediality, and ecology faced both by the people who made it and the audiences that view it.