Inscribing the Real

An ambiguity is built into the particular Chinese term for realism that gives this chapter its title.10 Xieshizhuyi 写实主义 combines a character meaning “to write,” “to inscribe,” or “to describe” (xie 写) with a character meaning “real” or “actual” (shi 实), followed by a suffix indicating an -ism (zhuyi 主义) or school of thought or art. Realism is thus literally a credo of inscribing the real, where credo can indicate anything from a set of techniques or conventions to a full-fledged aesthetic ideology. For Chinese artists, realism was intimately linked to modernity itself, in particular to the concepts of a modern nation and its citizenry. Chinese intellectuals of the early twentieth century—spurred by the historical rise and imperialist aggression of nation-states in the West, with their associated vernacular national cultures—believed that a modern, realist style had to replace the more stylized aesthetics of traditional Chinese literature, drama, and visual art.

By the early twentieth century, the modern practice of xieshi 写实, or “inscribing the real,” was (and still is) sometimes defined in opposition to what is considered the more traditional Chinese aesthetic of xieyi 写意, literally “inscribing the meaning/sense/idea/intention,” variously translated as expressionism, impressionism, conceptualism, and symbolism. In imperial China, xieyi referred specifically to a poetic, literati style of free-brushstroke painting, the opposite of which was not xieshi (as it would be in modern times) but rather the style of detailed court painting known as gongbi 工笔, or roughly “meticulous brush” painting.11 While gongbi referred to a specific genre of painting, the closest equivalent to a more general meaning of “realism” in premodern Chinese aesthetic theory may have been simply zhen 真, which signifies more broadly “authenticity” rather than necessarily a mimetically accurate representation of reality,12 so that a more formalist or expressionist work could just as well be credited with truthfulness, sincerity, or authenticity as a rigorously verisimilar one.13

In contrast to such a single-character sign for truth, authenticity, or an ultimate spiritual reality, the modern four-character term for realism, xieshizhuyi 写实主义, clearly implies a mimetic representation, promising access to an external reality through some form of inscription or recording. At the same time, the inclusion of that character for “writing” or “inscription” (xie 写) as well as the suffix for an -ism or doctrine (zhuyi 主义) suggests the potential for representation itself to become problematic. That is, if the real is “inscribed” by art through a doctrine or set of conventions, then a potential gap opens up between the representation and the reality it purportedly captures; the real real threatens always to elude or exceed the ostensibly realist work of art, and the process of inscription itself may take on autonomous qualities beyond the “reality” it records. Such autonomies of form over “real” content could range from the formalisms of high modernism to the generic conventions of popular entertainments. In either case, a tension develops between any claim to realist representation and the formal conventions that are employed to achieve such contact with the real.14

Dialectics of Realism and Convention

It is precisely this tension that Roman Jakobson explicates in his concise 1921 essay “On Realism in Art,” which sketches what can usefully be taken as an abstract model of the historical dialectics of realism. Jakobson begins by distinguishing two types of realism. One is “the conservative tendency to remain within the limits of a given artistic tradition, conceived as faithfulness to reality.”15 At the time of his writing, this tendency would have corresponded, in literature, to the nineteenth-century European realist novel or, in painting, to the kind of Renaissance-based “academic” realism still taught in established art academies. The opposing model of realism, however, was “the tendency to deform given artistic norms” and, moreover, to view such “deformation as a more accurate rendition of reality.”16 At the time, this tendency would have been embodied by the various modernist avant-gardes then erupting in both literature and the visual arts, a major purpose of which was precisely to eliminate the distance between art and life.17 In the former case of realism, “I hold my own values (the tradition to which I belong) to be the most realistic” and “I declare that the only genuine realism is the one on which I was brought up,” whereas in the latter case, “I can readily ascribe a realistic tendency . . . to forms which were never conceived as such” as long as they are shattering conventions while “a new truth [is] being sought.”18 Such an insight into what Jakobson calls “the extreme relativity of the concept of ‘realism’” suggests the potential for historical cycles in which a set of radically new approaches first challenges the old aesthetic order but then may itself become hegemonic and in need of challenging after its conventions no longer appear to provide access to the real but, on the contrary, seem to have become clichéd formulas that block such access—at which time the cycle can begin anew with the radical breaking of those (no longer) realist conventions.19

The focus for most of this book will be on historically situated, contextual readings of various claims to, or models of, realism in the history of mainland Chinese fiction film.20 How does any film, and the critical discourse surrounding it, stake out a claim to “realism” by reference to other types of cinema or other cultural forms that are deemed to be “unrealistic”? And how does such a claim to realism itself assert a set of conventions, whether preexisting or new at the time, that other (often later) artists or critics may in turn find artificial, constricting, and therefore in need of shattering for the sake of provoking a fresh confrontation with the real? Jakobson’s model suggests that such questions are inescapably questions of history, so that any claim to a universal, ahistorical definition of realism unwittingly finds itself caught in the same historical dynamic; if we were to settle on a singular concept of what constitutes realism, no matter how ingenious or true we might find it today, we ultimately may only be seen as adhering to another “conservative tendency” destined for historical oblivion.

Jakobson’s model of artistic change described in the preceding paragraphs, insofar as it shows that realism is inextricable from convention, could even be taken to support the idea that the ostensible referent of realism—the real itself, whatever that is—must be acknowledged as fundamentally inaccessible. Thus, in structural linguistics (developed by Ferdinand de Saussure and a major influence on Jakobson), signs are studied in terms of the signifier—the sound of a word, for example—and the signified, taken to be not a real thing but rather the concept that the word means, while the referent or nonlinguistic reality to which the latter may correspond is “bracketed” as beyond the scope of analysis of the system of symbolic representation.21 Such a potentially problematic gap between the “bracketed” real and the ways we experience, conceptualize, or represent it is of course evident in many forms of modern philosophy and critical theory—for example, Immanuel Kant’s idea of the noumenal “thing-in-itself”: while we have the idea that things exist entirely independently of our experience of them, they remain inaccessible to us in themselves.22

This can lead to a paradoxical view of the real as a mere effect of culture and therefore nothing but a kind of representational illusion; I only posit a reality “out there” because I am cued to do so by language or conventions “in here,” within my own culturally mediated experience. However, such a radical constructivism amounts to a form of philosophical idealism, neglecting the material realities of both natural and historical forces. An account of the pictorial arts in particular (and especially the photographic arts) that reduces all realism solely to conventions of representation—to the point that any role played by an “objective” external reality is denied altogether—commits an overtextualization of the image. The typical viewer does not merely experience a film, for example, as a set of symbolic signs to be read but also generally grasps the image (and, after the silent film era, sound) as an immediately recognizable perception—and does so not in a case of pure hallucination induced by artistic and ideological conventions but rather because, in the usual case of live-action photography, the images seen are in fact generated by a material reality that was before the camera. Thus a filmed shot, say, of a tree (the famous wind-blown leaves of Baby’s Meal [Repas de bébé; Louis Lumière, 1895], for example) has a level of cross-cultural recognition that no spoken sound or written word (“tree,” shù 树, etc.) could ever have, given the arbitrary differences between languages and the distinction between linguistic comprehension and perceptual apprehension. Even in the case of animation, the moving image provokes a visceral sense impression that is different from the reading of symbolic signs such as in a written text. I therefore reject any idea—possibly inspired by art historical studies like those of Erwin Panofsky and E. H. Gombrich, in addition to structuralism, poststructuralism, and the late twentieth-century “linguistic turn” in cultural theory and philosophy as a whole—that pictorial representation in general or photography in particular is a purely conventionalized and culturally constructed visual experience.23

Thus, despite the importance of Jakobson’s insight into the conventionality of realism, cinematic realism should not be analytically approached in a way that reduces realism entirely to the conventions of arbitrary symbolic systems. Even Jakobson’s essay implies that artists, particularly when they are trying to overcome entrenched conventions, are acting in some sense under the pressure of the real, in however mediated a form, so that his model suggests not just a historical dialectic between different sets of conventions—in which ones formerly labeled “realist” are periodically replaced by new ones in a seemingly random series—but implicitly an additional dialectic between art and reality itself. Art may always engage conventions of various sorts, but it cannot help but both impinge upon and be inflected by the real, if only as an “absent cause” that cannot necessarily be fully revealed or expressed in a narrative or work of art. Here I refer to Fredric Jameson’s rejection, even in the case of literature, of the poststructuralist notion that the “much maligned ‘referent’ itself” does not exist at all just because “it is inaccessible to us except in textual form, and that our approach to it and to the Real itself necessarily passes through its prior textualization, its narrativization in the political unconscious.”24 While it is no doubt true that we are always already within culture (and its “political unconscious”), we nonetheless do not merely navigate through arbitrary symbols or narratives in an endless game but also inevitably deal with the embodied pains and pleasures of our lived experiences of human history, including the broader biological and geological forces in which humans are embedded—experiences that help to forge and reshape those symbols or narratives in a never-ending, and never entirely successful, effort to account for the wonder and the agony of the real.

We must live, in addition and as a result, with an anxiety of the real—including an at least semiconscious awareness that there are aspects of the real that escape our current, culturally mediated construction of “reality.” As Catherine Belsey puts it, “if anything resists the sovereignty of the symbolic order, we always risk the uncanny possibility of an encounter that exceeds what culture permits us to define.”25 In this view, the real is not just some semiotic sleight of hand generated by culture, not just a constructed reality; it is, rather, the limit of culture, the determinate constraint and anxiety with which we contend and to which we adjust through culture. As Jacques Lacan precisely defined it, “the real is what does not depend on my idea of it,” and yet it is what our ideas and conventions seek to tame.26 What is at stake in theorizing realism is at least in part the fact that we always live in a certain limited reality that we provisionally understand and anticipate, yet we are never far from a limitless universe of things in time and space that we do not. In cinema, albeit in an always mediated fashion, the processes of a real universe press in on us, sometimes in reassuringly familiar forms, other times in ways novel and unsettling. Of course, audiences generally treasure the familiar—hence popular genre cinema—yet in cinema, we also risk an unexpected encounter with the real, an experience that can then change our “reality,” sometimes even to the extent that Jia Zhangke’s reality was changed by Yellow Earth.

Cinematic Realisms

This book distinguishes several different conceptual approaches to cinematic realism, each of which implicitly posits its own particular level of reality that is captured, whether it be the empirical material world; the world of embodied sensory experience; the illusory world of a fiction; a social or historical totality; or a philosophical, ethical, or ideological truth. Claims to realism in cinema have much overlap with similar issues in the other arts but also take on particular forms in assessing the medium of film. The categories sketched in what follows are meant to be neither exhaustive nor neat divisions; they are intended to categorize not necessarily films themselves but rather theoretical or analytical takes on realism—divergent approaches that often can be applied to the same films. Given the slipperiness and historical variability of the term realism, the differentiation of a few different models of it in the case of cinema can help to prevent category errors in which claims to one type of realism are mistakenly refuted by reference to assumptions rooted in a very different understanding of the term. What follows, then, is a preliminary mapping of six categories of cinematic realism—ontological realism, perceptual realism, fictional realism, social realism, prescriptive realism, and apophatic realism—with brief discussions that also point toward the most relevant later chapters in this book.

Ontological realism refers to the notion that the technology of photography provides a uniquely intimate and direct connection to material reality. This idea—which is sometimes (and somewhat problematically) summarized as “indexical” realism—arises from an important strain of classical film theory, represented in particular by André Bazin.27 The point is that photography—and, by extension, film, because conventionally, it was composed of twenty-four still photographs projected each second—has a uniquely strong tie to the objects it represents owing to the existential, causal relationship between object and photographic image. Traditional photography “automatically” records whatever is in front of the camera lens in a direct manner because light reflected from the object itself creates patterns in the emulsion on the filmstrip. Of course, human agency is required to decide—and quite possibly manipulate—what is to be filmed, but the optical and photochemical processes of the lens and filmstrip allow the camera to mechanically capture even parts of the material reality before the camera (“profilmic” reality) that perhaps were not intended or even noticed by the photographer.

This direct existential relation of the photograph to its objects appears to be what Bazin meant by his assertion that what photography gives us is “the object itself, but liberated from its temporal contingencies. . . . It has been created out of the ontology of the model. It is the model.”28 Many later film theorists also have drawn upon and elaborated similar arguments about the intrinsic ontological realism of photography or film—including, to name only a few, Dudley Andrew, Roland Barthes, Stanley Cavell, Mary Ann Doane, Laura Mulvey, D. N. Rodowick, and Philip Rosen.29

This idea of a direct relationship between the filmic image and a preexisting profilmic reality—which in some ways also subtends other arguments for realism listed in the following pages—is addressed most directly in chapters 1 and 7 of the current study. In chapter 1, which deals with discourses of realism and performance style in Chinese silent cinema and particularly the relationship of cinema to theater, we see how film’s ability to capture real locations as well as actors’ bodily performances was said to give it a privileged position among the performance arts in the context of an overall shift toward realism as China underwent semicolonial modernization. The relatively sudden rise in importance of realism in the arts in turn relates to a much broader discourse of scientism in Chinese intellectual life of the early twentieth century. Indeed, the type of real posited by ontological realism is in part that of the physical sciences—actual, discrete objects that make up “objective” material reality, the evidence of which can be recorded, measured, and at least in some sense known with the aid of instruments like the camera.

Chapter 7 reexamines claims to ontological realism in light of the rise of digital cinema, including both low-budget digital video (DV) filmmaking and special-effects blockbusters that make extensive use of computer-generated imagery (CGI). As has been widely discussed in cinema and new media theory, the digital revolution potentially calls into question the film medium’s previous claims to ontological realism, though I will argue that there remains a very important strain of Chinese digital cinema that continues to rely heavily on the viewer’s sense of the ontological realism of the image. The concept of photographic ontological realism is also key to several intervening chapters. For example, chapter 5 deals in part with how the ontological realism of cinema potentially lends increased credibility even to fictional representations of historical events, whereas chapter 6 notes that a reaction against master narratives of history may draw upon the capacity of film to register contingent detail and unexpected events.

Claims for what I am calling perceptual realism, our second category of cinematic realisms, emphasize not the ontological relationship between the cinematic image and a preexisting profilmic reality but rather the capacity of the moving image, whether photographic or otherwise (animation, CGI), to engage our perceptual apparatus and seize our attention, presenting itself to our senses as “real” whether or not it is seen as representing something else that is ontologically real.30 This could also be called phenomenological, experiential, or virtual realism, as it has to do with immediate visual and aural perception and the “real” embodied sensory experience that results even as the spectator remains intellectually aware that the images and sounds are not “real.” Extreme examples would include the feeling of dizziness induced by a shot from a camera on a moving roller coaster or the reflexive bodily flinching provoked by a sudden scare in a horror film. The difference between perceptual realism and ontological realism can be quickly grasped when we consider the fact that such reactions can be elicited by well-crafted animation or CGI as much as by live-action cinematography.

The tension between the perceptual realism of the moving image and the intellectual knowledge that it is “fake”—even live-action shots are really just flat images on a screen—always has been central to the fascination of cinema. Tom Gunning has described an “aesthetic of astonishment” among viewers of early film, who experienced a “pleasurable vacillation between belief and doubt” and appreciated the realism of the cinematic illusion all the more for knowing that it was an illusion.31 More recently Gunning has made further arguments regarding the centrality of movement to the perceptual realism of the moving image, with important consequences not only for early celluloid cinema but for animation and contemporary digital cinema.32 Indeed, perceptual realism is crucial to the concept of immediacy as elaborated by new media theorists Jay David Bolter and Richard Grusin, who note that most makers and consumers of visual culture “continue to assume that [the illusion of] unmediated presentation is the ultimate goal of visual representation and to believe that technological progress toward that goal is being made.”33 Many of the technological advances in the history of cinema—from the additions of sound and color to contemporary advances of 3-D and CGI techniques—have been driven by the goal of providing ever greater immediacy to the viewer’s perceptual experience, to move toward what Bazin called the “myth of total cinema”—a “complete realism” that provides a “perfect illusion of the outside world.”34

Of course, as Gunning emphasized, spectators, whether today or in early cinema, do not naively or delusively mistake an illusion for the real thing. Perceptual realism involves something far more complex: an ongoing, enjoyable play between realism and illusionism, between immediate sensory perception and intellectual awareness of the image’s artificiality. New technological developments—most recently virtual reality, for example—continually seek to renew that sense of wonder evoked by the tension between the sensory experience of perceptual realism and the cognitive awareness of illusionism.

Both ontological realism and perceptual realism, as defined herein, contribute to the more culturally dependent categories of fictional and social realism outlined in what follows, because the apparent reality and immediacy of the moving image help to facilitate audience absorption and promote the plausibility of a film’s fictional world. For example, the spectator’s possible mimetic response to the image—any tendency to imitate what is seen on screen, such as laughing or crying—is enabled partly by perceptual realism, and thus so is identification with the actor’s performance, as is discussed in the next chapter. However, the issues raised by perceptual realism will be returned to most directly with the discussion of CGI in chapter 7. We will see that perceptual realism can survive and even thrive when other, closely related types of realism (such as ontological and social realism) are abandoned.

In this book, fictional realism will refer to the tendency of narrative film to encourage diegetic immersion, in which a fictional world (diegesis) is taken as provisionally “real” for the purpose of feeling (positively or negatively) for characters and enjoying the story, whether or not the story’s world actually resembles our own. Alternative labels for this include “absorptive realism” and “diegetic illusionism.”35 So-called classical fiction film narration—as pioneered by figures like D. W. Griffith, standardized in Hollywood by the 1920s, adopted to varying degrees by other entertainment cinemas throughout the world, and analyzed in landmark works of film scholarship—can be thought of largely as an effective system for producing fictional realism.36 For example, the so-called continuity system of editing consists of conventions designed to keep the viewer focused on information from the fictional world rather than the formal devices that deliver that information. A type of edit like an eyeline match, for instance, habituates the viewer to connect a shot of a person looking off-screen with a subsequent shot of an object in a way that obscures the rupture of the cut by creating a larger unit of story-based meaning: character a looked at object b. The cut is thus made “invisible” to viewers insofar as they are focused on the events of the fictional world rather than the formal devices that reveal—or, rather, construct—such events.

The constructed world of a fiction film usually will have much overlap with the real world, but it also will have key differences. It may be set in a real city, for example, but have imaginary characters. The fictional world will tend to obey the basic laws of physics that govern our world—gravity, for example—except in cases where it does not, say, in a martial arts fantasy like Crouching Tiger, Hidden Dragon (Wohu canglong 卧虎藏龙; Ang Lee 李安, 2000).37 A fictional world may more or less successfully imitate our world in a multitude of its small details (costumes, furniture, architecture, incidental human behavior) precisely to draw us in and help us accept what might otherwise be the far-fetched make-believe of its story.38

V. F. Perkins has emphasized that we must take seriously the idea that a fiction film creates its own world, sharing aspects of ours but still distinct, which we are encouraged through cinematic technique to imagine as whole: “We are offered an assembly of bits and pieces from which to compose a world. Fragmentary representation yields an imagined solidity and extensiveness.”39 Anyone who has walked by a film shoot in progress likely has been struck by its artificiality. Almost any shot in a conventional fiction film would, if the camera were simply to track back farther than the shot allows, suddenly reveal many markers of its own constructedness—boom microphones, various types of artificial lights and reflectors, film crew members, possibly other cameras, dolly tracks, and so on. What is remarkable is the seeming ease with which a classical fiction film spurs us to lose ourselves in a fictional world that is in fact so fragmented and constructed on a shot-by-shot basis. Even the camera itself, while directing our gaze throughout the film, is disavowed by the world it shows us: “It is not that these characters are oblivious to the camera. There is no camera in their world.”40

Fictional realism thus results when film narration coaxes fragmentary representation to yield a cohesive world with the help of the viewer’s imagination. We tend to think that, for a successful film made according to classical norms, after we leave the theater, it will seem that it all fits together. Countless details will have been seen as appropriate to the film’s world; there will be no important hanging plot threads; gaps of time (narrative ellipses) and space (shifts of camera position between shots and scenes) will not have fundamentally disrupted the narrative chain of events; resolutions of major conflicts will have occurred; and in all likelihood (for Hollywood films at least), the coherent fictional world will be imagined to continue its existence in a provisionally positive and stable (if not necessarily happily-ever-after) mode after the film’s story is over (or at least until some new crisis necessitates a sequel).

Considering its function of making the film world seem like a unified whole, we can see why the film narration must rely on nondiegetic stylistic elements in addition to fictional events depicted on-screen. To return to the example of continuity editing, the potential disruption presented by the technique of the cut is trumped by the way it can facilitate the directing of audience attention to the most relevant characters, objects, and events in the story world. As long as editing remains roughly within the conventions of the continuity system, it enables classical narration and thus fictional realism; only when editing becomes especially unconventional does it draw attention to itself and potentially disrupt the viewer’s diegetic absorption. Another example is a typical fiction film’s soundtrack, which, in addition to the sounds that seem to emanate directly from the fictional world, frequently will include some nondiegetic music, which (Perkins points out) often “dictates a mood that is the film’s and not that of any of its characters”; it “is part of the film and not part of the fiction.”41 Soundtrack music, like continuity editing, though not itself part of the fictional world, nonetheless helps to unify that world into a seeming whole by presenting or reinforcing important information, even if that information is just a mood that we are encouraged to sense at a particular moment in the fictional narrative. Thus, in the classical film, elements of style and convention are more likely to facilitate fictional realism than they are to disrupt it.

As noted earlier, although here we distinguish fictional realism for analytical purposes, in our actual viewing experience, it is in many ways inseparable from ontological and perceptual realism, insofar as both the intrinsic realism of photography and the phenomenological realism of the moving image greatly facilitate our propensity to suspend our disbelief and lose ourselves in fictional worlds. However, fictional realism extends beyond those elements to include countless conventions of characterization and storytelling, relying on aspects like acting styles and genre standards to construct the story world. Generally, by the end of a classical Hollywood-style film, we feel we know all that is necessary to be satisfied that the world we have seen is unified and the story has reached, if not a final closure, then at least a “closure effect,” in which we are reassured that at least this story in the fictional world is over.42 However, Perkins reminds us that the classical self-contained solidity of the story world is only one option for a filmmaker and that it can be equally artistically valid to challenge such unity and closure: “The degree to which the filmmaker maintains the independence, solidity, and coherence of the fictional world is a matter of choice and a variable. The decisions have great significance for style and meaning but no immediate bearing on achievement.”43

This point brings us to one of the central concerns of chapter 2, which examines how late silent and early sound cinema in China freely borrowed conventions from classical Hollywood and elsewhere in the West yet also frequently disrupted the unifying function of the narration. Sudden shifts in characterization, tone, or theme might transform a romantic comedy into a work of social critique. Inserted metacinematic sequences might subtly call into question the escapism of entertainment cinema in general. Films might conclude with open-ended questions rather than any sort of reassuring closure effect. The resulting cinema of the Shanghai Left-Wing Film Movement, while greatly indebted to the classical standard and arguably constituting an example of its global spread as “vernacular modernism,” nonetheless in many cases departed fundamentally from classical norms, in part by eschewing the unity of classical fictional realism in favor of a more artistically avant-garde and socially critical practice.44

Chapter 3 as well is directly concerned with fictional realism, introducing in particular the degree to which a fictional film world presents us with what Umberto Eco has called an “epistemological metaphor.”45 Using three classic Chinese films of the late 1940s, it explores how a fictional world can model reality as either essentially knowable or unavoidably resistant to epistemological certainty. A film with a more classical form—including the canonical films of the Mao era (see chapters 4 and 5)—will imply both a totality (a unified, knowable fictional world) and a teleology (a purposive direction of events toward a predestined conclusion), while films with a more “open” form, such as many art films in general but also specifically some of the 1930s left-wing Chinese films discussed in chapter 2, the unique film by Fei Mu discussed in chapter 3, and the post–socialist realist cinema examined in chapter 6, will present built-in gaps in knowability, with no sure, final outcome. The former reassure us with the sense that the world is knowable, whereas the latter may uncomfortably confront us with the limits of our knowledge. The fictional world of a film thereby suggests a “structural homology” with epistemology in general.46

Insofar as fictional realism relates metaphorically to our own world, it also invites the question of plausibility, or whether the world presented by a film corresponds to what a particular community takes to be the “real” world. Indeed, in everyday discussions of the movies, the meaning of realism tends to revolve not around the ontological, perceptual, or fictional varieties discussed earlier but rather quite simply around the degree of correspondence between the fictional world and the social reality of a given audience. The effect of social realism builds on the three categories of realism already discussed; a film obviously relies quite heavily on the ontological realism of photography, the perceptual realism of the moving image and changing sounds, and the perceived cohesiveness or knowability of a fictional world to achieve a sense of verisimilitude. However, social realism goes beyond those aspects to embrace as well complex sociological and cultural factors. For example, do the language and behavior of the characters appear to be similar in their details to what spectators see in their daily lives? Do important categories of social life, such as gender, race, sexuality, nationality, and class, appear to be accurately represented? It is in factors such as these that cinematic realism also may be most apt to appear as historically and culturally contingent. Acting, for instance, that appears as entirely naturalistic at the time and place a film is made may appear as highly mannered to an audience a half century later or in a very different culture. The verisimilitude of acting styles is the focus of the next chapter, which considers how the shift from Chinese traditions of dramatic performance to modern notions about film performance was connected to historically specific cultural developments, such as the introduction of Western scientism into Chinese intellectual life, as mentioned previously.

Social realism is associated with the classic nineteenth-century realist novels of Europe, which continued to wield a strong influence on realist aesthetics well into the twentieth century, including in China. As early as 1854, Karl Marx glowingly praised “the present splendid brotherhood of fiction writers in England, whose graphic and eloquent pages have issued to the world more political and social truths than have been uttered by all the professional politicians, publicists and moralists put together.”47 What particularly impressed Marx was these writers’ damning portraits of the bourgeois middle class in England. Implicit in these observations from the mid-nineteenth century was a premise that would continue to guide proponents of social realism decades later, including many of the leading intellectuals of China’s New Culture Movement during the May Fourth era (approximately 1915–25) and beyond as well as important leftist European literary theorists, such as Georg Lukács: the idea that in unflinchingly and accurately reflecting society, complete with all its problems, realist art implicitly serves the function of social critique and thereby helps to lay the groundwork for social progress.

Owing to this function of critique, particularly in literary studies, social realism often is used interchangeably with critical realism. However, for the purposes of my typology of claims to cinematic realism, I would like to insist on the former category being in principle more expansive than the latter, which nonetheless forms a large and important subset within it. That is, in defining social realism as a sense of verisimilitude in the social reality constructed within a film, we do not necessarily have to take that depiction as serving the function of social critique in the service of some implicit reform or revolutionary agenda; if a spectator says “No Dai minority farmer from southern Yunnan would be speaking such perfect Beijing Mandarin!” she is applying the standards of social realism without regard to whether the film in question serves to critique society.

Still, the history of realism as a self-conscious practice in modern China, both in cinema and in the other arts, quite often has emphasized the potential of social realism for critique. The films of the 1930s Left-Wing Film Movement discussed in chapter 2 became canonized as classics of critical realism (although, as we will see, that is far from all they were). In revealing the desperate conditions of life for those suffering from both economic hardship and encroaching war during that decade, these films implicitly called for social reform, resistance, and even revolution. The tradition was renewed in the immediate postwar period, as chapter 3 examines. Afterward, with the founding of the People’s Republic of China (PRC) in 1949, the trend of politicized realism continued but also morphed from critical social realism into China’s version of socialist realism (see the subsequent discussion). Later, in the post-Mao reform era, the tradition of critical realism found renewed vitality with what I describe in chapter 6 as the phenomenon of post–socialist realism that spanned several decades.

In my typology of the various sorts of realist claims for cinema, I use the term prescriptive realism to refer to the idea that a film might seek to represent, not just reality as it now appears, but a truer reality that lies beneath the surface or is yet to be fully realized. Such a film thereby encourages the awareness or realization of that higher or deeper reality by presenting it as an abstracted form or ideal. In choosing the term prescriptive, I suggest an analogy with what is known in linguistics as prescriptive grammar—which aims to describe how people should talk according to the supposed inner logic of a language (or simply the “rules” as taught to children in school)—as opposed to descriptive grammar, which describes how people actually do talk (analogous to verisimilar social realism).

Chapter 1 shows how even mundane repetitions of Hollywood conventions can serve as prescriptive guides to the practice of being modern, and in later chapters, we will observe how social realism can abstract from surface appearances to try to show social reality’s deep structure, as Lukács as well as many Chinese critics argued. This is the moment when a realist mode begins to become prescriptive, insofar as a formal idea or argument of some kind is emphasized, potentially at the expense of surface-level verisimilitude. In the shades of difference between social realism and prescriptive realism, sometimes visible within a single film, we trace a historical progression of realism in China from the Republican to the Communist era (as well as from the revolutionary to the reform era, when the pendulum swung back).

In this book, then, prescriptive realism is most obviously represented by the cinema of the PRC during the Mao era (1949–76). As mentioned earlier, the cinema of that era could also be referred to as China’s variant on the international phenomenon of socialist realism, which was first formalized in the Soviet Union in the early 1930s and later spread throughout the Communist world during the Cold War. Commonly dismissed in the West as propaganda but in many cases achieving both popular appeal and artistic success, socialist realism was the official art form of the Comintern-based Communist bloc and was aimed explicitly at serving the revolutionary cause. In China itself, the term socialist realism was only in common use between 1953 and 1958; instead, other terms, such as proletarian realism, revolutionary realism, or even a combination of revolutionary realism and revolutionary romanticism, were favored in official discourse.

However, that variation in naming is only one of the reasons I use the more general (and less ideologically laden) term prescriptive realism. Chapters 4 and 5 show how the revolutionary realist form drew heavily on the conventions of melodrama, and indeed it shared a similar basis and function as melodrama does more generally—to use conventionalized character stereotypes and narrative scenarios to represent social truths and moral universes that underlie daily reality. In fact, I suggest in chapters 6 and 7 that this prescriptive mode of realism is not at all limited to socialist or revolutionary realism but also can describe, for example, the typical martial arts film—or, for that matter, a large number of American westerns, action films, or fantasy films. By the 1990s and 2000s, many Chinese films manifested prescriptive realism in the new form of capitalist realism—idealized representations of the neoliberal subject in the era of globalization. In each of these cases, “unrealistically” typified representations are used to convey “real” romantic, moral, or ideological ideals.

The final mode of realism that I define as a general type is apophatic realism, which returns us to the question raised earlier of a real that by definition escapes representation and thus calls for, as Laura Mulvey put it in a discussion of Abbas Kiarostami’s films, “an aesthetic of reality . . . in which the cinema acknowledges the limitations of representation.”48 Under the category of fictional realism, we touched upon Eco’s notion of the “open work,” in which a work of art serves as a metaphor for a more generalized epistemological uncertainty. We can relate this both to Jakobson’s discussion of the dynamics of realism and to the notion of a film’s “world” as analyzed by Perkins. Eco and Jakobson were responding, each in his own moment in the development of both semiotics and art, in large part to the challenges raised by the twentieth-century avant-gardes in visual art and literature. Jakobson noted that when an accepted set of “realist” conventions no longer appears sufficient to grasp a current reality, the urge to break those conventions will arise to facilitate the search for a new truth. Eco similarly attributes the rise of the “open work” to “a different vision of the world,” through which “an ordered world based on universally acknowledged laws is being replaced by a world based on ambiguity, both in the negative sense that directional centers are missing and in a positive sense, because values and dogma are constantly being placed in question.”49 Serving as an epistemological metaphor, a work of narrative art may thus manifest a general uncertainty or skepticism about accepted reality by opening up fissures in narrative conventions and gaps in the world thus constructed, leaving open questions, unresolved issues, and blind spots in the narration. As Perkins put it, in films, we are “led to understand more than we have seen”—that is, we grasp things that the narrative only implies; but at the same time, “to be in a world is to know the partiality of knowledge and the boundedness of vision—to be aware that there is always a bigger picture.”50 Indeed, there is always the possibility that the bigger picture will exceed what we are capable of seeing, and indeterminacy thereby becomes an inevitable, if generally denied, part of the world (any world). The occluded bigger picture actually is an unavoidable result of what Francesco Casetti calls “the partiality of the cinematic gaze”: “In showing one thing, it can and perhaps must leave something else out, even though what escapes us is sometimes at the heart of the story. Everything visible is accompanied by the invisible, and the invisible may constitute the essential.”51

To the extent that the real defies or exceeds a system of representation, it—or rather, its impossibility within the capacity of current conventions—can only be represented in a negative or subtractive manner, through silence, gaps, unsaying, that is, through apophasis.52 Just as apophatic theology insists that God or the absolute cannot be defined positively but only in terms of what it is not, apophatic realism is a mode of representation that seeks to acknowledge its own fundamental limitations, to build into its system an opening toward something beyond what representation can express.53 Though that opening may appear as a kind of nothing (gap, absence, lack, void) within what is represented, apophatic discourse is not nihilism but rather an acknowledgment that the real is more immeasurable, complex, rich, multiple, and mysterious than even a “realist” representation of it can possibly account for. Apophatic realism thus suggests that a proper ontology of cinema must concern itself not only with film’s capacity to reveal either known or new parts of reality to us but also with its capacity to point to what exceeds representation, to acknowledge the unrepresentable, the pure multiple of potentiality, an infinity that can only be figured as a nothing.

Of course, the idea, or anxiety, that something of the real inevitably escapes any human system of thought or representation has been expressed in many ways in various schools of philosophy and aesthetics in the modern West—for example, Ludwig Wittgenstein’s unspeakable, Martin Heidegger’s Being/Sein, Roland Barthes’s third or obtuse meaning, Jacques Derrida’s différance, Jacques Lacan’s Real, Luce Irigaray’s indeterminateness, Alain Badiou’s indiscernible, and Graham Harman’s fundamentally “withdrawn” objects, to name a few examples. The fact that these thinkers’ various projects are vastly different in so many ways only makes more notable the structural homology of a deeply apophatic moment in their thought—an insistence not only that there is something that exceeds representation or understanding but that that something may turn out to be the most important thing of all, threatening the system it exceeds but also potentially offering the utopian promise of liberation from that system. In fact, while seemingly characteristic of modernity, apophatic discourse goes back thousands of years in the West and also has a key position in the history of traditional Chinese thought and art, particularly in its Daoist and Chan Buddhist strains.54 Traditional Chinese thought and aesthetics may even lend themselves to the expression of modern apophatic anxieties, and we can trace their influences on certain Chinese filmmakers without necessarily resorting to simplistic cultural essentialisms.

Apophatic realism in film does not deny the powers of cinema to pursue the other types of realism already discussed, but it does find ways to gesture toward an openness of meaning or surplus of reality beyond the knowable world of a film. Cinema has many resources for indicating the unfigurable. Off-screen space can become not just an assumed extension of the story world or a space from which something new might appear but also a way to point to an irreducible limit to our understanding of the fictional world (which, again, functions as an epistemological metaphor for our own world). Edgar Morin described the camera lens as fixing a look “on the edge of the real”; off-screen space can be used to remind us of the real beyond that edge.55 Ellipses in narrated time can serve much the same function, particularly when they skip not only events that can be either inferred or ignored as irrelevant but ones that may be crucial to the unity of the story world. Bazin, discussing Italian Neorealist filmmaker Vittorio De Sica, notes that ellipses are essential to the overall structure of his narratives: “The empty gaps, the white spaces, the parts of the event that we are not given, are themselves of a concrete nature: stones which are missing from the building.” Bazin goes on to note that these gaps are themselves verisimilar, in that the lack of complete knowability of the story world mimics that of our own: “It is the same in life: we do not know everything that happens to others.”56 Besides being skipped by ellipses, events can be shown, but in ways that still highlight the necessary incompleteness of the spectator’s knowledge. We see the final event of Goodbye South, Goodbye (Nanguo zaijian, nanguo 南国再见, 南国; Hou Hsiao-hsien 侯孝贤, 1996), but from a vantage point that makes it impossible to really decipher. Alternatively, we may see the same events from a multiplicity of viewpoints in a way that calls any singular understanding of the events into question. Famously, in Akira Kurosawa’s Rashomon (1950), we see irreconcilable versions of ostensibly the same events, so that the apophatic realism of the film lies precisely in the unbridgeable gaps between the alternative versions of the story. More subtly, in Béla Tarr’s Sátántangó (1994) and Gus Van Sant’s Elephant (2003), we eventually circle around to see identical narrative moments from the perspectives of different characters, calling our attention to the fundamental discrepancies between different subjects’ experiences of the world and therefore the limits of each perspective. As a general category, apophatic realism can cover either a narration that is highly knowledgeable but less communicative—withholding information from the audience—or one whose knowledgeability itself is limited.57 That is, an apophatic technique could simply withhold information about the film’s world that it could have included, frustrating viewer expectations or knowledge of the fictional world, or it could more drastically suggest that there is something fundamentally withheld from knowledge or presence in our world in general; it could choose not to represent something that it could have, or it could gesture to something as fundamentally unrepresentable.

The idea of apophatic realism proposed here draws on various important strands in the history of film theory. Aside from Bazin’s thoughts on ellipses, a kind of apophasis is implicit in the idea of cinematic excess—the exceeding of narrative meaning or signifying exposition in a film—as explored in particular by Kristin Thompson in the case of fiction film and Bill Nichols in the case of documentary.58 In another context, Thompson and Bordwell have identified a radical openness of meaning in the films of Yasujiro Ozu as an aesthetically modernist impulse rather than, as others had suggested, showing the influence of Japanese Zen Buddhism, particularly the concept of mu (無; wu in Chinese) or nothingness.59 As suggested, however, such a dichotomy between Eastern “traditional” thought and Western modernism may be misleading when discussing apophatic realism in films from East Asia. Indeed, it is notable that our most in-depth discussions of that mode of realism will be in considering a filmmaker who has been called among the most distinctively “Chinese” in his aesthetics, Fei Mu (chapter 3).60

As mentioned earlier, these categories of cinematic realism are intended not as exhaustive or universal but mainly as heuristic devices for the purposes of the present study; other categories could be added for different objects of study.61 They are analytical tools to provide clarity in considering theoretical and critical claims, not necessarily a typology for categorizing specific films, nor even as always clearly distinguishable aspects of our moment-to-moment experience of watching films. In terms of actual spectatorial attention, for example, in a film with live-action photography and few special effects, ontological and perceptual realism will usually reinforce each other to the point of being essentially indistinguishable, while also contributing to a more general sense of fictional and potentially social realism. A strong feeling of fictional realism, that is, thorough immersion in a compelling story world, may in turn help to cultivate prescriptive realism—the sense that the fictional world provides a clarity that can mold our values and guide our actions in our own more corrupted and confusing world. The six categories I have named, then, are a reminder that realism can mean so many different things as to become meaningless unless such distinctions are made, yet each historically situated film or critical claim, when examined in all its complexity, likely will bring into play multiple approaches to realism.

Cinematic Conventions

André Bazin famously ends his essay “The Ontology of the Photographic Image”—a foundational text for the theory of ontological realism described earlier—with an enigmatic disclaimer: “then again, film is a language.”62 It is moments like this that have led later scholars, beginning with Peter Wollen, to graft Bazin’s arguments onto those of the pioneering nineteenth-century semiotician Charles Sanders Peirce, in particular his division of cultural signs into three types—iconic, indexical, and symbolic.63 In this argument, a film image not only looks like its referent in the manner of an iconic representation, such as a painting or drawing, but also has an indexical tie to it, pointing to the original by means of a direct existential relationship (the automatic inscription of light from the object on the filmstrip), thus constituting the medium’s ontological link to the real. Language, on the other hand, is a symbolic sign system, by which Peirce means that the relationship of sign to referent is a matter of arbitrary convention.

Thus Bazin’s statement that “film is [also] a language” puts us on notice that, besides drawing on the ontological/indexical realism of photography, cinema is in addition a matter of culturally constructed conventions. Indeed, as Thomas Elsaesser notes, another of Bazin’s landmark essays, “Evolution of the Language of Cinema,” expands on the language metaphor to understand cinema “as a practice governed by rules and conventions,” so that, while Bazin asserted ontological realism as “the content of the form” of cinema, he paradoxically also makes an argument “amenable to a constructivist view of verisimilitude.”64 Here Elsaesser cites another of Bazin’s paradoxical statements: “realism in the arts can only be achieved one way—through artifice.”65

Such observations point to the complexity of the dualism of realism and convention—terms that indicate not mutually exclusive opposites but rather a mutually constitutive pair. It is not simply that the intrinsic realism of photography, taking the “real world” as its raw material, is then manipulated by conventions of various sorts, from stylistic choices, such as framing and editing, to more elaborate genre conventions and cultural stereotypes. Rather, what strikes a particular audience as believably (or even shockingly) “realist,” in any of the varieties described in the previous section, inevitably is constituted largely by a set of methods that themselves may strike a different or later audience precisely as stylized or conventional—as we have already seen from the historical dialectics of realism suggested by Jakobson, in which realism is shown to be historically and culturally specific and thus highly relative and variable. The spectator’s perception of immediacy, transparency, verisimilitude, or truthfulness in a film is achieved only through the rigorous application of the technological possibilities of cinema at a particular moment as well as the skillful activation and manipulation of audience expectations formed through a culturally specific background, including previous viewing experiences.

We can begin to approach the question of cinematic convention through Stanley Cavell’s notion of automatism. This starting point might seem somewhat counterintuitive, given that Cavell’s initial conception of automatism appears to be closely related to Bazin’s ontological realism. Early in The World Viewed, Cavell describes automatism as cinema’s capacity for “removing the human agent from the task of reproduction”—a claim that appears very similar to Bazin’s observation that photography captures the world not just through an artist’s vision but through “another object” (the camera): “For the first time, an image of the outside world takes shape automatically, without creative human intervention, following a strict determinism.” Thus, “all art is founded on human agency, but in photography alone can we celebrate its absence.”66 Of course, neither Bazin nor Cavell is suggesting that there is no human agency in producing photography or film but only that a very important component of the power of those media comes not from the artist but from the “specific possibilities and necessities of the physical medium of film” or, more simply, from “film’s physical basis.”67 This basis includes both the material world that photography “automatically” captures and the “physical mechanisms of camera and projector,” which, at any given moment in the development of the medium, possess a particular set of determinate possibilities and limitations.68

Significantly, however, in a discussion of modernism in the arts in general and music in particular, Cavell expands the meaning of automatism to embrace not just basic physical and technological capacities but also the cultural forms through which the artist works:

The use of the word seems to me right for both the broad genres or forms in which an art organizes itself (e.g., the fugue, the dance forms, blues) and those local events or topoi around which a genre precipitates itself (e.g., modulations, inversions, cadences). In calling such things automatisms, I do not mean that they automatically ensure artistic success or depth, but that in mastering a tradition one masters a range of automatisms upon which the tradition maintains itself, and in deploying them one’s work is assured of a place in that tradition.69

This expansion of the meaning of automatism to include “form and genre and type and technique” combines (however “perverse” it may seem, as Cavell acknowledges) film’s ostensible ontological realism—its direct link to reality beyond human manipulation (though enabled by technology)—with the entirely human realm of cultural convention.70 As D. N. Rodowick summarizes, “automatisms in this sense are forms, conventions, or genres that arise creatively out of the existing materials and material conditions of given art practices. These in turn serve as potential materials or forms for future practices.”71 Or, as Cavell succinctly puts it, “a movie comes from other movies.”72 Conventions can function as automatisms in that they “act as variable limits to subjectivity and creative agency” as they “circumscribe practice, setting the conditions for creative agency and the artistic process.”73 Because the limits set by conventional automatisms are variable, artistic agency still is enabled—as we have seen, conventions can be altered or ignored in the “progressive” realist tendency identified by Jakobson, for example—but all artists cannot help but work within their own historical moments in the development of their art. An individual artist’s originality—prized by the standards of modernism but less important, say, for much traditional, commercial, or socialist realist art—cannot even be “identified without the background of repetitive automatism.”74 Even fiction filmmakers who radically break certain conventions are no doubt following others, perhaps even unconsciously, in that they generally, at least to some extent, are speaking the language they know—telling stories through film in the way that seems “natural” for their time and place and that can be understood by their intended audience.

The trope of film as a language certainly has its limits; indeed, the ontological and perceptual aspects of cinema’s realism themselves prove the inadequacy of the language metaphor.75 Still, insofar as film conventions function to some degree like a language, we can analytically group them into two types that are roughly analogous to syntax and semantics, or grammar and meaning. The first category includes technical facets, such as camera movement, editing, and countless choices in framing, focusing, lighting, and so on, whereas the more semantically oriented conventions have to do with the cultural meanings produced through settings, props, costumes, casting, performance, script, and so on—all of which adds up to stereotypes of genre, character, narrative scenario, and, ultimately, ideology. In practice, of course, these two broad categories of conventions are intimately intertwined. In “classical” film narration, the syntactical elements generally work to highlight the semantic message of the story, while in an avant-garde film, the opposite may well be true.

The most comprehensive theoretical treatment of semantic conventions in popular cinema to date is Jörg Schweinitz’s Film and Stereotype, which serves as a useful reference point for several later chapters in this book. Drawing on concepts from social psychology, linguistics, art history, and literary studies, in addition to film studies, Schweinitz defines stereotypes quite broadly, referring to them at one point as “conventionalized representational patterns shared and accepted by artists and viewers within a given period.”76 Like an idiom in language, a film convention or stereotype serves as a useful shortcut to meaning for a particular cultural community, but it also runs the danger of losing its aesthetic (and possibly communicative) power over time by becoming mere cliché. Relating to our earlier discussion, Schweinitz also refers to stereotypes as automatisms and their conventionalization as a form of automatization. Conventionalization is broadly but concisely defined as “standardization or codification in an intersubjective field,” with the implication that “stereotypes form and structure the intersubjective imaginary world of our time.”77 Representational tropes that may have begun with a conscious propositional meaning can become unquestioned automatisms that sustain an ideological worldview for a public without regard to whether their original perceived truth is still consciously evaluated. Indeed, the process of conventionalization tends eventually to be accompanied by what Schweinitz calls derealization, through which film stereotypes gradually lose their originally claimed connection with reality. For example, character types “first seeming to be plausible representations of reality increasingly resemble puppets in a game as they become entities of convention.”78 Even later, such stereotypes can undergo a kind of “secondary semantization” in which they represent the entire context in which they emerged, such as a particular historical social sphere or cultural genre.79 This in part illuminates how conventions that seemed comfortably “realist” at one moment—not just fitting into but indeed helping to construct the “reality” of their time—may strike later viewers as highly stylized or artificial, pointing not to the “real” world but to a particular cultural formation that subsequently has become passé or even alien.80

One example that we will trace in later chapters is Chinese revolutionary cinema. That tradition began with a critical realist impulse that coexisted, sometimes awkwardly, with conventions of Hollywood entertainment cinema in classic Chinese films of the 1930s (chapter 2). It continued with postwar masterpieces like Crows and Sparrows (Wuya yu maque 乌鸦与麻雀; Zheng Junli 郑君里, 1949), which seemed to grasp and represent, through such means as character types, the precise social totality of its own historical moment of political revolution (chapter 3). In cinema of the Seventeen Years (1949–66), partly under the influence of Soviet socialist realism, Chinese revolutionary cinema developed an arsenal of codified stereotypes that (as is the case with classical Hollywood genres, such as the western) both entertained mass audiences and reinforced conventionalized and affectively charged collective values regarding morality, behavior, politics, and community belonging (chapter 4). The eventual sedimentation of such codes into clichés coincided with the bureaucratization of political rule in general and the lessening of the ideological fervor of revolution. This danger was answered politically by the Cultural Revolution (1966–76), which in cinema eventually resulted in a striking new phenomenon, the films based on the revolutionary “model works” of opera and ballet. This new cinematic form, however, was even more rigorously conventionalized than the popular cinema that preceded it, so that, particularly given the model works’ repetitive appearances in culture (not just in cinema and drama but also through magazines, books, posters, and so on), its stereotypes became easily “denaturalized,” or “conspicuous” as conventions (chapter 5).81

By the beginning of the post-Mao period, in a case of the “secondary semantization” described by Schweinitz, the mere citation of one of the model works could serve to represent the milieu of the Cultural Revolution as a whole. Indeed, whereas on one hand, a reaction against the conventions of revolutionary or socialist realism led to the phenomenon of post–socialist realism—an “aesthetics of opposition” in Schweinitz’s terms82 that I argue is a long-term trend across multiple “generations” in postrevolutionary Chinese cinema (chapter 6)—on the other hand, to this day, the stereotypes of Mao-era revolutionary cinema continue to be reflexively cited in contemporary culture for purposes ranging from parody to nostalgia.

As the example of revolutionary cinema’s realism and convention illustrates, this book offers a narrative of Chinese film history with interconnections between the various chapters, though without pretending to be seamless or comprehensive. My interests are both historical, seeking to recover particular moments in film history in their complexity, and theoretical, using examples from the Chinese experience to explore more general questions of cinematic realism and convention. Chapter 1, “Acting Real in Chinese Silent Cinema,” uses the debates over performance styles in silent cinema and the effort to set cinema apart from stage drama as means of introducing some of the basic concepts of realist film theory, while also exploring how realism, with its close connection to broader notions of science and objectivity, became a master signifier of modern aesthetics across the arts in early twentieth-century China. Chapter 2, “Shanghaiing Hollywood in the 1930s,” explores how the conventions of classical Hollywood cinema were redeployed in Shanghai’s early sound cinema to create a left-wing cinema that was not only politically progressive but also formally radical, leading to a reconsideration of the relationship between classical Hollywood and Chinese cinema as “vernacular modernism(s).” Chapter 3, “Realism and Event in Postwar Chinese Cinema,” examines three classic films of the late 1940s in terms of their temporal structures and relations to history, whether lingering on the irretrievable losses of the recent war, circulating around a void of impossible desire that seems to suspend time, or hurtling forward through a historical rupture. Chapter 4, “Prescriptive Realism in Revolutionary Cinema of the Seventeen Years,” reconceptualizes Mao-era cinema as a form of prescriptive realism and reconsiders the idea of sublimation in revolutionary aesthetics via an examination of classical narration and genre stereotypes. Chapter 5, “Socialist Formalism and the End(s) of Revolutionary Cinema,” extends these considerations into the model drama films of the Cultural Revolution, the “socialist formalism” of which exposes contradictions between dialectical materialist philosophy and state Communism that the Cultural Revolution ultimately failed to resolve. Chapter 6, “A Long Take on Post–Socialist Realism,” proposes that a transgenerational “post–socialist realism” threaded through the decades of the post-Mao period in part as a reaction against the prescriptive socialist realism that preceded it. Chapter 7, “Chinese Cinematic Realism(s) in the Digital Age,” addresses the question of how different modes of Chinese film have been transformed (or not) by the transition to digital cinema, from CGI-heavy blockbusters to independent films made on DV.

In the end, the question of realism concerns not only the correspondence between a fictional world and our world but, more fundamentally, the relation between appearance and being as such. The relationship between the fictional world and our world revolves around questions like “verisimilitude” and “plausibility”; the relationship between appearance and being, however, suggests another, even more basic problem of realism. Here, by appearance I mean not simply a visual image but more broadly the way our world or reality appears to us precisely as structured through codes and representations, including language, the encyclopedic knowledge of our culture, the stereotyped images and conventions that help us understand our daily lives (including the fictional worlds we consume), the deep-rooted assumptions that guide our thought and behavior, and (not least) the sensory and cognitive capacities of our biology. In this sense, our “world” serves as a filter through which we (fail to) grasp the multiplicity of being; rather than forcing us to confront the real, it is more likely to help shield us from the profound indeterminacy that is our lot, given the limited senses and intelligence with which our species faces the universe (or multiverse) in which we exist.

Because appearance, or being as grasped—a coherently perceived image or, more broadly, a “world” that it fits into—inevitably falls short of being as such, when we humbly acknowledge such limits, we may embrace a realism (what I am calling apophatic realism) that deals not only with what the fictional world tells us about our world but with how the narrated world potentially gestures to the limits of appearance, pointing to the traces of new possibilities, challenging our sedimented sense of ourselves and our world, and placing us—temporarily, exhilaratingly—outside our usual framework. As Dudley Andrew summarizes a strain of film thought and practice that stretches from Bazin and Italian Neorealism through the other Cahiers critics and French New Wave and on to the contemporary cinema of Jia Zhangke and Hou Hsiao-hsien, such a realism both insists on cinema’s intimate tie to the real and renounces mere representation as the nature of that tie.83 This cinema exists not primarily to entertain, instruct, or reassure us by elaborating our existing understanding of reality but to jolt us out of that reality through an encounter with the real.