I.
Poetics of Cinema
Sometimes our routines seem transparent, and we forget that they have a history. It's commonplace for academics in the humanities to assume that every field consists of objects of study and diverse "methods" for studying them. In literary studies, you have texts, and you try to understand them by applying a variety of doctrines about literature or language or life. You maybe a phenomenologist or a Lacanian, a follower of deconstruction or poststructuralism or cultural studies, but everybody, explicitly or unawares, subscribes to some method.
Familiar as it is, this way of thinking isn't eternal. It emerged only 60 years ago, out of the boom in college literary criticism that followed World War II. The benchmarks are Stanley Edgar Hyman's The Armed Vision (1948) and René Wellek and Austin Warren's Theory of Literature (1949). Both books set forth the novel idea that literary studies played host to distinct "methods."1 Intrinsic and extrinsic; textual and contextual; sociological, Marxist, psychoanalytic, and archetypal: The categories invoked by Hyman, Wellek, and Warren have a distinctly modern ring. Thereafter, book-length studies explored one method or another, applied to this or that author, and editors compiled anthologies pitting one method against another for the sake of classroom instruction. With the emphasis on "practical criticism," the professor could take a poem or play and train upon it the guns of competing methods. The anthology-of-approaches genre became a going concern in the 1950s and 1960s, and it continues to flourish.2
Film studies quickly subscribed to critical Methodism. One of the most popular anthologies, Movies and Methods, first appeared in 1976.3 You can argue that this tactic helped give media research a path into the university. Even if somebody thought that the object of study lacked importance—what intellectual would study
11
12
Poetics of Cinema
Poetics of Cinema
13
Hollywood?—-a set of up-to-date approaches constituted intellectual bona fides. But it seems to me that film studies accepted, too unquestioningly, the literary humanities' conception of method itself.
In film studies, as in its literary counterpart, "method" comes down to meaning "interpretive school." An interpretive school, I take it, asks the writer to master a semantic field informed by particular theoretical concepts and then to note certain features of films that fit that field. The writer then mounts an argument that relates features of a film to the theory by citing the film, quoting from relevant theorists, and creating associative links between the semantic field and the film. For example, to a psychoanalytic critic, certain semantic features enjoy a particular saliency: semantic oppositions, like male and female or sadism and masochism, along with concepts like the deployment of power around sexual difference. The critic will then pick out textual cues that can bear the weight of the semantic features, such as the narrative roles assigned to men and to women, or the representation of the act of looking. The critic will then mount an argument, perhaps using the rhetoric of demystification, to show the significance of the semantic projections, from field to text, that the critic generated. Every recognized "method"—phenomenological, feminist, Marxist, or whatever—follows something like this routine. They all aim to produce interpretations, which I take to be ascriptions of implicit or symptomatic meanings to texts.4
Poetics is a somewhat different enterprise. It doesn't constitute a distinct critical school, so it isn't parallel to any of the doctrinally defined methods. It has no privileged semantic field, no core of procedures for interpreting textual features, and no unique rhetorical tactics. Although interpretations don't lie outside its province, the status of interpretation isn't quite what it is in the doctrine-driven approaches. Put another way, the domination of methods-based thinking has yielded various hermeneutics, but poetics is something else again.
The Tradition
Aristotle's fragmentary lecture notes, the Poetics, addressed what we nowadays recognize as drama and literature. Since his day, we have had Stravinsky's Poetics of Music, Todorov's Poetics of Prose, a study of the poetics of architecture, and of course the Russian formalists' Poetics of the Cinema? Such extensions of the concept are " plausible, because it need not be restricted to any particular medium. Poetics derives from the Greek word poiesis, or active making. The poetics of any artistic medium studies the finished work as the result of a process of construction—a process that includes a craft component (such as rules of thumb), the more general principles according to which the work is composed, and its functions, effects, and uses. Any inquiry into the fundamental principles by which artifacts in any representational medium are constructed, and the effects that flow from those principles, can fall within the domain of poetics.
Some further distinctions are useful. A research project in poetics may be pri-marily analytical studying particular devices across a range of works or in a single work. You can, for instance, study assonance and alliteration in sonnet lines. Or the
1
project can be predominantly theoretical, laying out conditions for a genre or class of work. Examples would be Aristotle's account of tragedy and Gerard Genette's map of how temporal relations can be represented in narrative* There is also historical poetics, the effort to understand how artworks assume certain forms within a period or across periods. Usually, any project will involve all three perspectives, but one or another will predominate. For example, although Aristotle recognizes the changes that tragedy underwent over several centuries, he concentrates on building a theory of the genre as a more or less ideal type.
Along another dimension, a project in poetics may be predominantly descriptive, outlining the principles of "making" without preference for one option or another. Alternatively, it may be more prescriptive, favoring certain options. Several versions of poetics, most often those mounted by artists themselves, are prescriptive. The neoclassicists of eighteenth-century English literature promoted a poetics of reason and generality, obliging the artist to propound universal truths. The poet, in Dr. Johnson's phrase, "does not number the streaks of the tulip."7 Romantic theorists of a later generation argued for an alternative poetics, one that finds beauty in unique forms. Coleridge finds himself entranced by the frost on a windowpane because it has crystallized into traceries suggesting trees or seaweed.8
All these distinctions may seem rather abstract, so let me illustrate the diversity of poetics, as 1 conceive it, with two exemplary essays, one about literature, and the other about cinema. The poet W. H. Auden confessed that he enjoyed reading criticism that could "throw light upon the process of artistic 'making,'" so it's not surprising that several of his literary essays approach authors or genres from the perspective of poetics.9 In "The Guilty Vicarage," he provides a compact account of the classic detective story. His emphasis is predominantly theoretical, finding the crux of the genre in an Aristotelian pattern of action: A murder is committed, many are suspected, and the killer is revealed and punished. This permits Auden to distinguish the detective story from an adjacent genre such as the suspense tale, in which the murderer's guilt is known from the outset. From the core plot action, he deduces several other features of the genre. The concealment of the murderer's identity raises problems of narration (key information must be withheld from the reader) and the structuring of time (the action usually needs to unfold in a short span, before the killer can escape). The crime must be murder, because the absoluteness of the act forces society to act on behalf of the victim. The plot demands a certain community (typically, a closed one) and setting, along with particular roles assigned to victim and murderer, the innocent participants, and the detective. The thematic crux of the genre is that of an Edenic society that is ruptured. Murder creates a crisis because it "reveals that some member has fallen and is no longer in a state of grace."10 Auden concludes by considering the role of the reader in enjoying the tale's recovery of innocence, whereby the guilty party is revealed to be radically different from the reader. Auden's anatomy of the genre doesn't trace historical conventions or analyze a single story in detail, and it is shot through with judgments about what is preferable in the genre, such as obedience to the classical unities of time, place, and action.
14
Poetics of Cinema
Poetics of Cinema
15
André Bazin's classic essay "The Evolution of the Language of Cinema" illustrates a more historical approach to poetics.11 Bazin argues that Western film style is not best considered as a development from silent cinema to sound cinema, but rather as a process in which two tendencies, present at the start, collide or blend. On one hand, there is the tendency toward recording reality; on the other, there is the urge to abstract from reality, to create artifice. The earliest films relied on recording, but in the mature silent cinema of the 1920s, this tendency became a minority option. In various ways, D, W. Griffith, the German expressionists, the French avant-garde, and the Soviet montage filmmakers exemplify the triumph of stylization, making cinema a vehicle for abstract concepts and formal experiment. But the coming of sound, with its unremitting tie to recording speech, banished the formal trend and led to a middle way represented by classical studio style. The sound cinema, for Bazin, marked the decline of antirealistic filmmaking and the emergence of a relatively realistic, moderately manipulative style relying on analytical cutting, shot/reverse shot, and other features that do only a little violence to the event in front of the camera. But at the same time, directors like William Wyler and Orson Welles recovered the recording capacity of the camera, with long takes and deep-space staging presenting the event in all its completeness. Still, having learned the lessons of classical cutting, Wyler and Welles organized their images so that they were more articulated than the primitive frames of the Lumiěres and Georges Méliěs. Wyler and Welles discovered how to present a scene in a single shot but retain all the changes of emphasis to be found in an analytical breakdown into several shots. Citizen Kane (1941) and The Little Foxes (1941) constitute "a dialectical step forward in the evolution of film language," a powerful reconciliation of opposed tendencies.
Throughout, Bazin relies upon theoretical distinctions, such as intershot effects versus intrashot effects, types of montage, distortion versus fidelity, spatiotemporal unity versus discontinuity, and shallow space versus depth. Note that these aren't like the semantic fields governing doctrine-driven interpretive methods. Bazin, quite reasonably, holds his concepts to be principles determining the stylistic construction of any film whatsoever. The analyst can correlate the choice of devices with intended effects. So Bazin proposes that the highly visible montage in Sergei Eisenstein's films yields a step-by-step layout of meaning, whereas Welles' and Wyler's depth of field conveys dramatic information through the simultaneous presence of various elements.
While Auden is interested in plot and theme, Bazin focuses on style. Auden is almost wholly unconcerned with the history of the detective story, treating it as an ideal type, but Bazin puts his theoretical categories into motion, letting them measure the development of Western cinema from primitive filmmaking to neorealism. Like Auden, however, he can be prescriptive. It's plain that he favors cinematic styles that, preserve the spatial and temporal integrity of reality, so the history he traces carries a note of triumph. Welles and Wyler have ushered in a Hegelian synthesis of all previous cinema.
One can quarrel with Auden's and Bazin's essays,12 but they illustrate some of the options available to the poetician. They also exemplify the possibilities of a poetics that is grounded in both rational and empirical inquiry. Each writer's categories
are explicit and open to criticism on conceptual grounds. Just as important, each writer invokes concrete evidence that allows us to appraise his claims. This appeal to empirical evidence, or "facts," does not make poetics into an "empiricism," at least in any interesting sense of that term. A poetics can be rationalist or empiricist, Kantian or phenomenological, deductivist or inductivist, idealist or positivist. Whatever its ontology or epistemology or discovery procedures, a poetics appeals to intersubjec-tively available data that are in principle amenable to alternative explanation.
To a great extent, an exercise in poetics typically takes as its object a body of conventions—genre conventions in "The Guilty Vicarage," and stylistic conventions in the "Evolution of the Language of Cinema." Conventions, in film as in other domains, lie at the intersection of conceptual distinctions and social customs. Auden characterizes the detective story as at once an adventure in reasoning and a social ritual of casting out the sinner. Bazin's realist aesthetic leads him to range stylistic devices along a continuum whereby some are less "conventional" than others. Nevertheless, in studying patterns of editing and mise-en-scene, he is invoking a structured set of options that are quite salient in Western cinema. These options constitute norms, a central concept in poetics that is explicitly signaled by our two writers.
The two sample essays exemplify still other aspects of the line of inquiry I'm exploring. They show that we might join observation of general tendencies with a scrutiny of particulars, such as Auden's discussion of various detective heroes and Bazin's account of influential films and directors. The writers consider both "texts" and "contexts": Auden situates the detective story within Christian societies, while Bazin looks at trends of genre and narrative that affect style. And both poeticians presume that the artwork results from choices within a craft tradition. The fictional detective may be propelled by purely intellectual motives, as Sherlock Holmes is, or he may be seeking to redeem a soul, as Father Brown is. Likewise, Bazin's case rests upon the possibility that Welles could, if he'd wished, have shot and cut Citizen Kane in the manner of It Happened One Night (1934). Craft practices always offer a range of options, and the choices made by the artist will be correlated with some purpose—the design of the work or an effect on the perceiver.
My initial questions and my exposition of the Auden and Bazin essays should raise several questions about how this approach works. What, for example, is the status of the "principles" studied by poetics? I'd argue that the principles should be conceived as underlying concepts, constitutive or regulative, governing the sorts of material that can be used in a film and the possible ways in which it can be formed. At what level of generality are these principles pitched? The degree of generality will depend upon the questions asked and the phenomena to be studied. If you want to know what makes Hollywood narratives cohere, "personalized causality" may suffice as one constructive principle; if you want to know what distinguishes a film noir from a musical, that principle isn't up to the job. For some poeticians, principles are held to be laws on the model of covering laws in physical science, but we needn't push that far. You could assert that a concept—say, Bazin's distinction between what happens within shots and what happens between shots—is foundational, but that the ways in which
16
Poetics of Cinema
Poetics of Cinema
17
filmmakers use that as a constructive principle vary so much that we can trace only general tendencies.
Are the principles conceived as "specific" to cinema in some sense? Although certain poeticians have assumed a distinction between the cinematic and the non-cinematic, this view isn't a postulate of poetics as such. You can assume that any film could be studied by poetics, with no film laying any closer to the essence of the medium than others. You could, though, also argue that the distinction between cinematic and noncinematic is not a substantive but rather a functional one, to be filled out in different periods with different content. Or you could use the distinction in an explanation by seeking to show that in particular circumstances, this pair of concepts entered into the norms of filmmaking practice because filmmakers believed in some version of it.
Poetics is often assumed to aim merely at descriptions or classifications, so I should elaborate a little on the range of explanations it offers. There's no need to assume any one model of causation and change. Bazin argues for a broad dialectic through which cinema evolves toward an ever more faithful capturing of phenomenal reality. This is a teleological explanation. One could also propose an intentionalist model that centers on filmmakers* localized acts of choice and avoidance. Then there's the possibility of a functionalist model of explanation, whereby the institutional dynamics of filmmaking set up constraints and preferred options that fulfill overall systemic norms.13
Nor need poetics be confined to "immanent" explanations that refuse to leave the field of cinema, art, or representational media. Nothing in principle prevents the poetician from arguing that economics, ideology, cultural forces, or inherent social or psychological dispositions operate as causes of constructional devices or effects. There is likewise no need to cast poetics as offering "scientific" explanations (although, again, some poeticians have done so). Poetics has the explanatory value of any empirical undertaking, which always involves a degree of tentativeness about conclusions. On the other hand, one shouldn't dismiss historical research's affinities with science too quickly, because there are many scientific disciplines, such as geography and archaeology, which fall short of predictive accuracy but have good records of ex post facto explanatory power.'4 It's probably best to say that poetics joins the overarching tradition of rational and empirical inquiry to which science and kindred disciplines belong.
Finally, and to return to a difference with the doctrine-driven methods of film studies, explanation in poetics doesn't confine itself to issues of what films mean. Of course, meaning in one (very general) sense comprises a big part of what poetics describes, analyzes, and explains. But meaning in the narrower sense that is the product of film interpretation (a "reading") isn't necessarily the goal of the poetician. Films produce many effects, ranging from perceptual ones (why certain color schemes dominate films of a particular period) to conceptual ones (how we know that X is the protagonist), and these matters film interpretation never seeks to elucidate. Historical poetics, in particular, tends to offer explanations rather than explications. Still, critics are makers too, and we could analyze their materials, principles, and
concerns with effect. That is, we can try to explain interpretations. Finally, practical criticism, focused on particular films, can be fruitfully informed by poetics. Several of the essays that follow try to show how.
Domains and Tendencies
Traditional poetics in any medium distinguishes among three objects of study: thematics, large-scale form, and stylistks. Thematics considers subject matter and theme as components of the constructive process. The researcher may study motifs, iconography, and themes as materials, as constructive principles, or as effects of constructive principles. Auden does this with Christian themes in "The Guilty Vicarage." Similarly, film scholars have revealed how genres present recurring imagery, myths, and motifs, whereas other writers, inspired by art-historical research, have shown the importance of iconography in popular cinema.15
Taken broadly, thematics informs many people's thinking about cinema. Scholars and journalists commonly scan movies for social stereotypes and political attitudes. The stock phrase "representation of race, class, and gender" invokes themes in a loose sense. Yet although studies of these matters can be enlightening, they wouldn't usually constitute a contribution to poetics, because they're often not concerned to link themes to constructive principles. They don't typically show how the overall design of the movie, including areas not obviously related to the stereotypes on display, requires the stereotypes in order to achieve its particular purpose or effect. Nor do they treat the themes discovered as part of historical traditions of art making.
For example, the films of Ozu Yasujiro often evoke the theme of the transience of human life. This theme is common in the world's art, and it's especially prominent in Japanese poetic traditions. It was reworked in Japanese popular culture early in the .20th century, when a newly modernizing Tokyo was seen to embody the ephemeral-ity of existence: Yesterday's building would be demolished and replaced by something up-to-date tomorrow. The theme is highlighted in Ozu's films, when characters talk about past pleasures and look forward to a moment in the future when they will recall what they're doing now. The theme also finds visual expression in Ozu's use of conventional imagery of transience like clouds, smoke, and streetlights switched off or on. Often items of setting, like laundry on a line or household utensils, disappear or change position from scene to scene, teasing us to recall an earlier moment in the film. Ozu carries the theme of constant change, itself a cultural cliche, from broad social sources into the intimate drama and down into the details of filmic texture.16
Sometimes there's a tension between thematic givens and the film's overall design. Early in Laura (1944), a sharp opposition is set up between McPherson, the rough-edged police detective, and Waldo Lydecker, the effete bon vivant who has been the patron of the murdered Laura. McPherson is a real man, and Waldo a sissy. Before she died, Laura threw over Waldo for Jacoby, who is at once a handsome, athletic man and a sensitive artist. In any realistically motivated plot, Jacoby would be a prime suspect. Remarkably, however, McPherson doesn't investigate him at all. It seems likely that Jacoby, being a blend of the extremes that define the major male characters, is too
18
Poetics of Cinema
Poetics of Cinema
19
much of a threat to the final purpose of the plot, the romantic union of McPherson and Laura. So the plot is distorted to make the themes fit.17
Thematics lies perhaps closest to method-dominated criticism, but within the tradition of poetics there have been wide-ranging theoretical and methodological debates of a kind not seen in interpretive approaches. What are themes, and where might we find them? Should we consider them unique particulars (e.g., Romeo), common motifs (lovers), generalizations (love), or semantic oppositions (romantic love versus social obligation)? Is a theme a priori, something that the artist inherits and reiterates, or is it post facto, something that perceivers create in order to endow the artwork with coherence and significance? For many people, the discovery of themes is a major reason for engaging with artworks; for some poeticians, the discovery of themes has little bearing on interpretation as usually conceived. In the 1960s, two Soviet researchers created a "poetics of expressiveness" that treated themes as "deep structures" undergoing various transformations before being concretized as surface patterns of the text.18 Here, Yuri Shcheglov explains, the theme is not a message or separable content that the reader carries away, but rather a principle provided by the analyst in order to account for the formal features of the text.19 Somewhat similar was the effort of Michael Riffaterre to show that a poem is generated from an underlying verbal formula, often a linguistic cliche. The text elaborates this donnee without necessarily naming it. Thus a Cocteau poem elaborates a traditional motif, "the inn of death," through imagery of travel, dying swans, and other figures of speech.20
A historical poetics of cinema is likely to consider themes as given materials that are transformed by traditions of form and style. Soviet films of the 1920s and 1930s were charged with representing the emergence of "revolutionary consciousness," and filmmakers like Eisenstein and V. I. Pudovkin could take these themes for granted and explore increasingly allusive ways to signal them. The prepackaged nature of Leninist ideology helped directors in the montage style and Socialist Realism create oblique, flamboyant ways of saying what everybody already "knew."21 It isn't only openly rhetorical cinema that can be illuminated by the idea of theme as a cultural given. Noel Carroll has pointed out that many narrative films can be seen as illustrated homilies. They presuppose vague commonplaces in order to be intelligible, as Back to the Future (1985) assumes that anything can be altered by individual striving.22 By studying commonplaces in circulation in a given period's culture, we can often link cinema to other media and social life.
A second domain of poetics is that of large-scale form. The poetics of literature explores principles of progression and development governing the well-made play, the sonnet, or the adventure novel. We students of cinema lack a term for those transmedia architectonic principles that govern the shape and dynamics of an entire film. The most prominent research domain here is the theory and analysis of narrative, which is a fundamental constructive principle in films.23 In this book I've devoted a separate essay to narrative form and several essays to particular films or traditions, so here I'll just mention that there are other compositional principles that poetics should investigate. A film can be organized as a rhetorical argument, or it can collect an array of categories, as in a catalogue. The form may be associational, as in
a film lyric like-Stan Brakhage's Scenes From Under Childhood (1967-1970), or based wholly on abstract similarities and differences (Ballet Mecanique, 1924).34 In practice these types can combine, as when films like Fahrenheit 9/11 (2004) recruit narrative patterns to a larger rhetorical impulse. R Adams Sitney's classic book Visionary Film proposes another taxonomy of large-scale forms, as manifested in the postwar American avant-garde.25
Stylistics, the third leg of the poetics tripod, deals with the materials and patterning of the medium as components of the constructive process. Bazin's "Evolution" essay is a model of stylistic history of cinema. Most of the stylistic studies in this book concentrate on visual patterning—staging, shot scale, composition, editing, and camera movement—but that doesn't mean that sound doesn't matter. Scholars with better-trained ears than mine have studied how techniques of sound recording and reproduction shape the stylistic texture of a film, or how the score contributes to the overall stylistic dynamics.26 On this last front, many scholars have refined our understanding of film music; the number of studies of film music far exceeds the number of studies of cinematography, editing, or other techniques of the image track.
Philosophers of art have long debated how to define the concept of style. Expressive theories treat style as the manifestation of artistic personality or emotional states, rhetorical theories treat style as a matter of impact on the audience, and objective theories consider that it consists of objective properties of the artwork's formal design. Then there are conceptions of period style, national style, and the like. All of these ideas have proven fruitful for researchers studying the poetics of the arts.
One Poetics of Film
Given the great variety of research programs within the broad domains of poetics, I'm going to use the rest of this essay to characterize the threads of reasoning that wind through the pieces in this book. I propose a version of poetics that rests upon film analysis. Fór mé, the most interesting questions grow out of particular films. This angle of approach can draw inspiration from rich traditions in adjacent fields. Art historians such as Heinrich Wólfflin, Alois Riegl, Erwin Panofsky, and E. H. Gombrich show us how to systematically track forms and styles in the visual arts and explain their changes causally.27 In literary theory, the Russian formalists and the Prague structuralists-most notably, Viktor Shklovsky, Yuri Tynianov, Boris Eikhenbaum, Jan Mukařovský, and Roman Jakobson—proposed both concrete analyses of literary works and larger explanations for how they functioned in historical contexts.28 More recent literary theorists, particularly Meir Sternberg, have also provoked me to formulate a position on narrative in cinema. Leonard Meyer, Charles Rosen, and other musicologists have likewise furnished models for thinking about form and style in relation to historical change. The essays in this book are explicitly indebted to these thinkers.
Most academic books about cinema carry at least a dollop of theory, so it's best to be clear about the role of theory in the essays that follow. I sometimes draw upon film theorists of the pre-1970 period, such as André Bazin, the Soviet filmmakers of the 1920s, and Noel Burch, but that doesn't entail that I'm committed to Bazin's
20
Poetics of Cinema
Poetics of Cinema
21
phenomenology or Eisenstein's odd version of materialism or the early Burch's serialist theory of film. If we take a film theory to consist of a set of propositions explaining the fundamental nature and function of all cinematic phenomena, the poetics I'm setting out doesn't amount to a theory in that sense. It's best described as a set of assumptions, a heuristic perspective, and a way of asking questions.29 It's frankly empirical and tries to discover facts and truths about films.
When I first floated the idea of a historical poetics of cinema, I anticipated objections to an empirical program of this sort. After all, many said in the 1980s, there aren't facts but only "facts" (that is, social constructs that vary according to time and place), and there can be no truth (usually identified, erroneously, with absolute truth or final truth or capital-T Truth). Since the rise of cultural studies, an area of inquiry that wants to discover ways in which audiences appropriate films, I don't have to be so defensive. Film scholars have come to realize that any descriptive or explanatory project is committed to some grounding in intersubjective data. All intellectual disciplines seek to find out how things are.
Of course, there's no question of letting facts speak for themselves. We can't discover plausible answers to questions about films' construction without carefully devising analytical concepts appropriate to these questions. But not all concepts are equally precise, coherent, or pertinent, and so we may evaluate competing conceptual schemes. Crucially, we're not complete prisoners of our conceptual schemes. We may specify our ideas in an open-textured way, so that exceptions leap to our notice. The poetics I'm proposing makes claims that are theoretically defined, open-ended, corrigible, and potentially falsifiable.
This is a direct result of its not being a general theory of film. If I'm bent on substantiating the belief that every film constructs an ongoing process of "subject positioning" for the spectator, nothing Ifindinafilmwill disconfirm it. Given the roomy interpretive procedures of film criticism, I can treat every cut or camera movement, every line of dialogue or piece of character behavior, as a reinforcement of subject positioning. The theory becomes vacuous, because any theory that explains every phenomenon by the same mechanism explains nothing. On the other hand, I can ask how Hollywood films secure unity among successive scenes, and answer with something more concrete-say, that one scene often ends with an unresolved causal chain that is soon resolved in the following scene. Here I've said something that's informative. It isn't self-evident, it isn't discoverable by deduction from a set of premises, and it's fruitful, leading to further questions. Does this constructive principle suggest some hypotheses about the nature of narrative norms in Hollywood? (It does.) Do films in other filmmaking traditions utilize more self-contained episodes? (They seem to.) Most important, the answer I supply could be disconfirmed. If it is disconfirmed, I need to rethink the data and, indeed, the question itself. Shklovsky's counsel of skepticism should be our guide: "If the facts destroy the theory—so much the better for the theory. It is created by us, not entrusted to us for safekeeping."30 Ideally, our hypotheses are grounded in a theoretical activity rather than a fixed theory. This activity moves across various levels of generality and deploys various concepts and bodies of evidence. It seeks to be driven
by data and midrange concepts rather than by abstract or absolute doctrines, and it can be recast or rejected in the light of further investigation.
In being question centered and focused on particular phenomena, the poetics I envision somewhat resembles the practices of inquiry in scientific endeavor. Stephen Jay Gould writes,
Progress in science, paradoxically by the layman's criterion, often demands that we back away from cosmic questions of greatest scope (anyone with half a brain can formulate "big" questions in his armchair, so why heap kudos on such a pleasant and pedestrian activity?). Great scientists have an instinct for the fruitful and the doable, particularly for smaller questions that lead on and eventually transform the grand issues from speculation to action____Great theories
must sink a huge anchor in details.31
Likewise, here is Francois Jacob, discoverer of RNA's function as a "message molecule" for transmitting genetic information:
The beginning of modern science can be dated from the time when such general questions as "How was the Universe created? What is the essence of life?" were replaced by more modern questions like "How does a stone fall? How does water flow in a tube?" While asking general questions led to very limited answers, asking limited questions turned out to provide more and more general answers.32
Someone will object that this appetite for midsize questions may suit the hard sciences, but studying culture and history can't be so precise. Yet C. Wright Mills, no positivist by any description, suggests that the "sociological imagination" is characterized by a middle way.
Classic social science, in brief, neither "builds up" from microscopic study nor
---------"deduces down" from conceptual elaboration. Its practitioners try to build and
to deduce at the same time, in the same process of study, and to do so by means of adequate formulations and re-formulations of problems and their adequate solutions. To practice such a policy... is to take up substantive problems on the historical level of reality; to state these problems in terms appropriate to them; and then, no matter how high the flight of theory, no matter how painstaking the crawl among detail, in the end of each completed act of study, to state the solution in the macroscopic terms of the problem.,.. Controversy over different views of "methodology" and "theory" is properly carried on in close and continuous relation with substantive problems.33
Mills' duality echoes the two options we're usually offered in the humanities. You may tackle very tightly focused projects, which supposedly lead to steadily accumulating knowledge; you can't make bricks without straw, as they say. Alternatively, you do Grand Theory, where you can't make a move without getting all your abstract doctrines correct beforehand. Elsewhere I've advocated that film scholars could pitch a project at a middle level, asking questions of some scope without deep commitments
22
Poetics of Cinema
to broad doctrines, and using the answers to those questions to build hypotheses of greater generality.34
I quote these worthies at length not to show that poetics is a science. My point is that as compared with Grand Theory, it aims at satisfying general demands of rational and empirical inquiry. Take, for example, the notion of norms. In the essays that follow, I assume that it's often useful to ask how a film relates to sets of transtextual norms. These operate at various levels of generality and possess various degrees of coherence. For instance, in most studio-made narrative films, the credits sequence characteristically occurs before the first scene, but it may also, as lesser options, occur after a "precredits sequence" or during the first scene. Such norms, although "codified" are not reducible to codes in the semiotic sense, because there is no fixed meaning attached to one choice rather than the other. Jean-Luc Godard's decision, in Ditective (1985), to scatter the credits sporadically through the first 14 minutes yields rather unusual effects on our apprehension of the story, but no definite meanings automatically proceed from it.
For a long time, people training to be composers or performers studied "music theory" and aspiring painters studied "art theory." These terms didn't refer to inquiries into the nature and functions of their respective art forms, still less to the Grand Theory that permeates the humanities today. Music theory was about how to write counterpoint, how to orchestrate effectively, or how to build a symphony up from phrases. Art theory was about composition, color values, and the like. Music theory and art theory were repositories of craft knowledge, stated in more or less principled fashion and invoking the proven success of inherited norms. To a large extent, poetics is a systematic inquiry into the presuppositions of artistic traditions. It's a practice-based theory of art. We want to know the filmmakers' secrets, especially those they don't know they know.
Craft norms are historically variable; the music theory taught in Paris conservatories in the 1880s was very different from that taught to musicians in India. That's why we need a historically inflected poetics, one that recognizes that art is made differently under different circumstances. A historical poetics, it seems to me, should also be alert for commonalities among apparently diverse norms. Conventions shared across distinct art traditions can be as important as those of narrower provenance, as I try to show in the essay following this.
A poetics can reveal both change and continuity among norms by reconstructing a historical context. How does this work fit into a tradition? How does it repeat, revise, or reject its forerunners? This sort of thinking is commonplace in mature disciplines. Consider, for instance, Boris Eikhenbaum's essay on the stories of O. Henry. Here Eikhenbaum traces changes in the writer's oeuvre against the background of the history of the American short story and its masters, Washington Irving, Poe, Harte, and Twain. O. Henry's work, he claims, displays a series of formal experiments moving from cyclical construction to apsychological characterization and reflexive parody. He discusses how the writer ironizes the sentimental style that was then dominating mass literature. He shows causal connections between O. Henry's innovations and changes in American literary tastes and magazine publishing. He concludes with the
Poetics of Cinema
23
claim that O. Henry at the end of his life looked forward to writing something more straightforward about manners and morals, thereby looking ahead to the loosely plotted, slice-of-life form characteristic of Dreiser and Sherwood Anderson. "The O. Henry story, with parody at its core, opened the way for this regeneration."35 I'm sure that literary historians could dispute Eikhenbaum's claims, but his essay shows how one can fruitfully blend formal analysis with a coherent conception of historical continuity and change. It's a model of a midlevel research project.
Finally, a few words about form, formalism, and "formalism." Given the intellectual lineage I've claimed, and the fact that in the 1980s the approach I sketch here came to be called neoformalism, I should try to block possible misunderstandings. Sometimes "formalism" implies an art-for-art's-sake position. But if that view implies that artworks don't have consequences for morality, behavior, and society, I don't hold it. Some people use the term to indicate that poetics considers only "form" and not "content," or "culture " or whatever other subjects the critic thinks more important. More accurately, the poetics I propose looks at artistic form as an organizing principle that works not on "content" but rather on materials: not just physical stuff like film stock or the items set before the camera but also themes, subjects, received forms, and styles. Out of these materials, the relevant principles create a whole that aims to achieve effects. By studying form in the sense I mean here, we can understand how cinema turns materials circulating in the culture into significant experiences for viewers.
Pretty obviously, those experiences both shape and are shaped by a variety of cultural forces. Several essays that follow indicate that by studying cinematic form in the three domains already indicated, the poetics I'm proposing need not cut off cinema from larger dynamics of social life. True, my questions bear more upon the "how" of film than the "what," but both are necessary for full understanding. I hope that any reader of "Three Dimensions of Film Narrative," "Convention, Construction, and Cinematic Vision," or "Who Blinked First?" will see that I do engage with matters of culture, though my conception of it may not correspond completely with that promoted by practitioners of cultural studies.
Poetics: A Program
As I conceive it, a poetics of cinema aims to produce reliable knowledge by pursuing questions within two principal areas of inquiry. First is what we might call analytical poetics. What are the principles according to which films are constructed and through which they achieve particular effects? Second, there's historical poetics, which asks, How and why have these principles arisen and changed in particular empirical circumstances? In my view, poetics is characterized by the phenomena it studies (films' constructional principles and effects) and the questions it asks about those phenomena—their constitution, functions, purposes, and historical manifestations.
This research program doesn't put at the forefront of its activities phenomena such as the economic patterns of film distribution, the growth of the teenage audience in the 1950s, or the ideology of private property. We may need to investigate
24
Poetics of Cinema
Poetics of Cinema
25
such matters, but they become relevant only in the light of the questions about filmic construction that guide the inquiry. Underlying this hierarchy of significance is the assumption that, although in our world everything is connected to everything else, we can produce fresh and precise knowledge only by making distinctions among core questions, peripheral questions, and irrelevant questions.
I can specify further. At the risk of seeming cute, I can characterize the research framework I propose by six P-words: particulars, patterns, purposes, principles, practices, and processing. These are related, so that by examining any one of them, we're likely to find connections to others.
Putting a film or set of films at the center of our concern can lead us toward the most atomic items we can detect. Perhaps we're struck by a line of dialogue, or a certain cut, or a moment in a performance, or an unusual sound. Details are always worth noticing, and they're often what critics point to in justifying an appreciation or an interpretation of a film. Some critics, from the surrealists to the present, have made a virtue out of celebrating the isolated particulars as ends in themselves. Details can give us a buzz.36 Sometimes, though, the items that seize our attention seem puzzling. Why, for instance, do characters in film stare at each other so constantly and intently, and why do they blink so seldom? Despite theorists' interest in The Look and The Gaze, eye behavior of this sort hasn't attracted a lot of attention, but it's an intriguing feature of filmic storytelling. So the particulars that attract our attention can seem either unique to the film or something, perhaps even something trivial, that it shares with other films.
Any poetics goes beyond particulars. The items that we notice belong to patterns. The hero's single wisecrack belongs to a stream of comments that he makes, a lot of them wisecracks. Low-key lighting may become associated with a certain character or locale. Most of the essays that follow take patterns of narrative or visual style as a primary object of inquiry. But there's a problem here. Any element that we spot can be situated within an indefinitely large number of patterns. What makes some patterns salient, either in the act of watching the movie or in the course of our analyzing it?
Our best candidate is the purpose that we can assign to the pattern. Our hero utters wisecracks because characterizing him this way fulfills some functions in the story. Perhaps his insolence gets him into trouble with his boss, or makes him appealing to us, or serves as a foil to a more phlegmatic character, or all of the above. The way characters quite unrealistically stare at each other in films has particular functions, as I try to show in a later essay. Sometimes filmmakers will acknowledge the purposes that their strategies fulfill, but more often we have to posit some plausible ones ourselves. And it goes without saying that anything we pick out may be serving many functions, and several devices may be working in harmony to achieve one overall purpose.
Recall a device in what's come to be called the classical Hollywood tradition. We might notice that in Nick of Time (1995), the climax depends upon a race against the clock. Watch a few more thrillers, and you find the same thing. But then you watch some romantic comedies or domestic dramas, and you find that often they too focus their resolution on the pressure of time. We have a conventional pattern of action, one that today's screenwriters call the "ticking clock." This pattern in turn functions to
increase dramatic tension, leading us to expect that the dramatic issues will come to a definite climax. It's significant as well that fiction films in other traditions, such as Antonioni's L'Avventura (1960) and Hou Hsiao-hsien's The Puppetmaster (1993), don't rely on deadlines to bring the action to a resolution. These narratives are governed by other purposes.
Often it's useful to conceive the artist's purpose in terms of problems and solutions. At a mundane level, a filmmaker wants to achieve some pattern or effect. Something blocks this, so the filmmaker contrives a way to achieve the effect he or she wants. The result may turn out to be more complicated than what was initially planned. The essay on CinemaScope in this volume argues that in facing technological constraints, some filmmakers returned to a form of staging that was thought to be outmoded, with intriguing results. Or an adventurous filmmaker may actually court problems, laying down self-imposed constraints in order to stimulate her or his imagination. Both Ozu and Mizoguchi Kenji did this, the one refusing dissolves and usually situating his camera always at a lower height than what he photographed, and the other favoring long takes and framing the action from a high angle. Each director's narrowing of the artistic bandwidth not only yielded surprising expressive resources within the field he staked out, but also forced him to deal with a cascade of new problems that a simple technical choice brought in its wake.37 We should therefore remember that functional explanations can sometimes make things too tidy. Every decision is a trade-off, yielding not only benefits but also costs.
When we find repeated items, patterns, and functions across several films, we can ask about the principles underlying these factors. Most often those principles will be in the nature of norms, those explicit or implicit guidelines that shape creative action. I've already suggested that conventions are central subjects for poetics, and we can think of norms as the principles that govern conventions. Some norms operate at the small scale, whereas others shape the formal design of whole films. Sometimes norms are formulated as crisp rules, but most often they are rules of thumb and operate in the background, learned and applied without explanation or even awareness. (Filmmakers know a great deal more about their activity than they articulate.) We're often left to infer the relevant norms by noting regularities and then seeking out evidence that could count for or against.
Take, for instance, the staging techniques that emerge in feature filmmaking in Europe during the 1910s. For several decades, most film historians were content to call this tradition "theatrical" because it relied on lengthy takes of action recorded at some distance,Ja Although the historians recognized that the editing innovations of Americans (chiefly Griffith) were highly patterned, the European films didn't seem organized to the same degree. The historians could support their claims by the fact that several critics and filmmakers writing in the 1910s and 1920s praised films that displayed close-ups, alternating editing, and accelerated cutting. But once stylistic historians began to look closely at the European films of the period, the films no longer seemed backward. It was evident that the films were patterned to provide clean, clear uptake of story events within an integrated space. The patterns turned out to be governed by deep-space staging and, more fundamentally, by the perspectival space
26
Poetics of Cinema
Figure 1.1 In Victor Sjostrom's Ingeborg Holm (1913), the clerk in the family's store slips perfume to a pretty customer as Ingeborg comes in from the back room. The camera position lets us see that the cash register blocks the clerk's theft from Ingeborg's view, an action that wouldn't be evident from every seat in a stage theater; indeed, from some sightlines, the register would conceal Ingeborg's arrival from the spectator.
Figure 1.2 Ingeborg Holm: Later, a similar camera angle shows Ingeborg's crisis. The bill collector presses her at the cash register, while the guilty clerk pops into a patch of space behind them. Again, this shot wouldn't work on the stage because the clerk's expression wouldn't be visible from most seats. Sjostrom realized that cinematic space, unlike theatrical space, is resolutely monocular; only the camera's eye matters.
provided by the camera lens. These principles are in significant ways wntheatrical. Filmmakers used both two-dimensional composition and three-dimensional blocking to guide the spectator's attention to the unfolding story in ways impossible to achieve on the stage (Figures 1.1-1.2). Once the system of norms became apparent, one could go back to documents and find records by filmmakers that pointed to their self-conscious awareness of the pyramidal playing space of cinema.39 Those statements had been available to historians for decades, but they sprang into relief only after close viewing revealed that there could be alternative norms.
Some norms are probably quite local, such as the deadline-driven climax characteristic of Hollywood cinema. Other norms apply to a surprisingly wide variety of. films. We expect that mainstream filmmakers will tend to place the chief action at the center of the frame, but an otherwise transgressive movie like Jeanne Dielman, 23 Quai du Commerce, 1080 Bruxelles (1976) has recourse to the same tactic (Figure 1.3). Across the world, filmmakers of the 1980s and 1990s tended to build scenes out of "singles"—shots, usually quite close, of individual characters—rather than ensemble framings. But norms are systematic and hierarchical, so any dominant principle will mold others to its needs. In the Viennese classical style of Western music, Charles Rosen argues, the centrality of the articulated phrase shaped rhythm, texture, and dynamics.40 Similarly, once a filmmaker accepts the norm of centered composition, she or he will tend to adjust staging procedures, lighting, color choices, and editing patterns accordingly. The reliance on singles in modern cinema forces filmmakers to cut more frequently, in order to trace the flow of the conversation and to remind the viewer what characters are present.41
Poetics of Cinema
27
Figure 1.3 Jeanne Dielman, 23 Quai du Commerce, 1080 Bruxelles (1976): When Jeanne Dielman fails to button her coat properly, traditional centering procedures make sure we notice this as a sign of her breakdown.
As my examples indicate, studying norms is an exercise in extrapolation. By trying to chart the range of constructional options open to filmmakers at various historical moments, we come up with results that are always open to revision. In practical research terms, attention to historically changing devices, patterns, principles, functions, and norms moves us beyond the single film to groups of films. By positing alternative norms, our work becomes comparative in a rewarding way. Instead of the couplet norm-deviation, we can posit competing systems of principles, operating at roughly the same level of generality. We find varying norms of narration and style in Hollywood cinema, "art cinema," Soviet montage cinema, and other modes.43 In Hollywood cinema, for instance, the norm of cogent storytelling favors not only a ticking clock but also a coordination of that with other conventions, such as causal continuity and a duplex plotline involving both work goals and romantic goals. Although it may be momentarily helpful to characterize art cinema narration as a "deviation" from Hollywood principles, it's more enlightening to characterize it positively, as possessing its own fairly coherent set of storytelling principles (as I try to do in a later essay in this book). Recognizing that we're engaged in a comparative exercise allows us to give equal weight to one norm and another.
Moreover, we don't have to postulate every historical change as a deviation from a norm. I've already suggested that we can often think of changes as driven by problems, some inherited from tradition, others devised by the filmmaker. There are as well many ways to realize norms, some obvious, some subtle. The most striking stylistic changes in film history often don't stem from absolute innovation but rather from a recasting of received devices. Welles' deep-focus staging in Citizen Kane is a famous instance,43 but we could say much the same of Godard's cutting in Breathless (1960), which recasts orthodox continuity principles (matching on movement and eyelines) into new patterns, to new effect. An innovation isn't necessarily a deviation. I suggest in a following essay on Robert Reiner t that his rather odd-looking films are the result of taking to a limit certain staging principles that governed
28
Poetics of Cinema
mainstream European cinema of the 1910s. Reinert, we might say, broke with the norm by carrying it to extremes.
Some people think that studying norms necessarily celebrates them. That's not the case. Norms can be studied objectively without marginalizing alternatives as freakish or unacceptable. In fact, studying what's norm driven can make us sensitive to what runs athwart the norm. And, again, what is most fruitful isn't just celebrating the demolition of standard principles but also the putting forth of alternative systems. Kristin Thompson, for instance, has been concerned to demonstrate how the works of Eisenstein, Jacques Tati, Godard, Jean Renoir, and others provide not fitful deviations from norms but rather systematic innovations in thematic, stylistic, and narrative construction,44 We can balance a concern for revealing the tacit conventions governing the ordinary film with a keen interest in the unusual film that, subtly or flagrantly, challenges them. Accordingly, new concepts will often have to be forged. To account for Ozu's editing, Thompson and I had to devise the concept of the "graphic match" and to spell out how Ozu's across-the-line shot/reverse shots do not willfully transgress rules but rather achieve particular functions within a larger, idiosyncratic system of 360-degree space.45
The aims and principles we detect in films are rooted in activities. Filmmakers work with tools and materials, operating within institutions that offer both constraints and opportunities. These factors can be summed up under the rubric of practices. How shall we understand these practices? Two ideas can guide us. First, there is a rational agent model of creativity. This follows from the idea that the filmmaker selects among constructional options or creates new choices. Our task becomes that of reconstructing, on the basis of whatever historical data one can find, the creative situation that the filmmaker confronts. The rationality at stake is largely one of means-end reasoning. Assuming a certain end in view, certain options are more likely to fulfill it than others. If you want to raise tension at the end of a film, then it's not unreasonable to add a deadline, especially if the tradition in which you're working offers you several ways to indicate that deadline's approach (including shots of ticking clocks). Filmmakers have reflected, to various degrees of detail, upon their creative choices, and this literature offers a rich legacy of insights into practices.46
This isn't to say that the filmmaker becomes the sole source of the film's construction and effects. A second, institutional dimension of practice forms the horizon of what is permitted and encouraged at particular moments. The filmmaker works, most proximately, within a social and economic system of production, and this involves tacit aesthetic assumptions, some division of labor, and standard ways of using technology. When we want to mount causal accounts of some features or forms, the mode of filmmaking practice is a good place to look.47 It's not just that the filmmaker's choices are constrained; they are also actively constituted in large part by socially structured factors of this sort. In the Hollywood studio system of the 1920s and 1930s, for instance, the continuity script not only became a way to rationalize production. It also encouraged workers to think of a film as an assemblage out of discrete bits (shots, scenes), and the individual filmmaker found choices and opportunities structured accordingly.48
Poetics of Cinema
29
In most sorts of filmmaking, practices are crystallized in routine ways of doing things. These form the filmmakers' craft. Filmmakers of the 1910s laid down sticks or chalk marks to show their players the pyramidal space that the camera was taking in, and they marked out a " front line" that signaled the point of closest focus. Screenwriters today often plot their films around a three-act structure; if you don't have a turning point about 25 minutes in, you're flouting standard practice. Certain choices of lenses or film stocks come to be preferred, for aesthetic, institutional, and practical reasons. Hollywood directors are expected to shoot a great deal of coverage, piling up alternative camera setups that offer a great deal of choice in the final editing. "The financiers want every conceivable option," notes director John Madden. "They want you to shoot a wide shot, they do not want you to cover it in one shot, they do not want you to say, 'I don't need that reverse.'"49 However strong tradition may be, though, filmmakers still have choices about how they will utilize the options available to them. Working with shared technology and traditions, directors of early CinemaScope films differentiated themselves. Hong Kong filmmakers of the 1980s did the same.
By invoking means-ends reasoning and the options offered by tradition, I don't want to suggest that filmmakers brood deeply on every decision, that there's no spontaneity or flash of insight guiding their creative choices. When we want to mount a causal account, starting with assumptions about rationality and craft practices can serve as a methodological default. All other things being equal, and as a point of departure, it's fruitful to assume that a filmmaker makes choices in order to achieve some purposes, as those might be defined by the tradition in which the filmmaker works. Moreover, the rational agent model doesn't rule out lucky accidents or flights of inspiration. Where creative ideas come from is fairly mysterious, but once the artist has the idea, she will make choices about how to integrate it into the work at hand, and these will be inflected by means and ends, purposes and patterns. A burial scene in Red River (1948) is enhanced by a cloud passing over the assemblage. The filmmakers . took advantage of the wayward cloud because it helped fulfill the purpose they had in view for the scene.
Treating norms and craft practices as traditions also implies that some continuity underlies changes we might observe. In this respect, I suppose, the historical side of poetics is conservative, often trying to remind people that things that seem brand-new almost always proceed from longer-lived conventions. If we call Crash (2005) and Happy Endings (2005) "hyperlink narratives," we're implying that their formal principles are pretty new, arising from recent technological changes like the Internet.50 But when we look closely, those principles are revealed to be modifications (sometimes slight ones) of norms that have been used in cinema for decades. In a later essay, where I call such movies "network narratives," I argue that social networks, which have always been with us, have recently become salient for reasons having to do with culture and pressures within the film industry. The films rely on long-standing traditions of multiple-protagonist plotting, making it possible for audiences to track the action easily. The motto of historical poetics might be that of Shakespeare's Lear: Nothing comes of nothing.
30
Poetics of Cinema
Poetics of Cinema
An Excursion on Reflections and Zeitgeists
All my talk of conventions and practices and individuals acting within institutions runs afoul of some long-standing intuitions. For many educated people, the most important questions about cinema revolve around its relation to culture. The persistence of this concern is itself puzzling to me. In no other domain of inquiry I know, from the history of science and engineering to the history of music, literature, and visual art, is there such unremitting insistence that every significant research project must shed light on society. Scholars can freely study iambic pentameter, baroque perspective, and the discovery of DNA without feeling obliged to make vast claims about culture's impact on said subjects. Is cinema important and valuable solely as a barometer of broad-scale social changes?
In any event, many will suggest that the framework sketched here seems oblivious to the ways in which films reflect their cultures. How can I forget that social anxieties, economic crises, and cultural tensions govern the form and content of movies? Remarkably, nearly everybody believes this. The idea shapes the Sunday New York Times think piece about how the movies of the last few months capture the current Zeitgeist. It informs the belief that we can define periods in American popular art by presidential eras—leave It to Beaver as cozy Eisenhower suburban fantasy, Forrest Gump (1994) as an expression of Clintonian post-Cold-War isolationism.51 Reflectionism may be the last refuge of journalists writing to deadline, but it also underlies a great deal of what academics pursue under the rubric of cultural studies. That mass entertainment somehow reflects its society is, I believe, the One Big Idea that every intellectual has about popular culture. Yet there are good reasons to be skeptical of it.
It's commonly felt that cinema, being a popular art, tends to embody the attitudes or emotions of the millions of people living in a society. Yet this argument needs shoring up, because it easily becomes circular. (All popular films reflect social attitudes. How do we know what the social attitudes are? Just look at the films!) We need independent and pretty broadly based evidence to show that some deep needs of the audience exist and are being addressed by a film. Just because Spiderman (2002) was a huge success doesn't automatically mean that it offers us access to America's national mood or hidden anxieties. People spend time with a piece of mass art for many reasons: to kill an idle hour, to meet with friends, to find out what all the fuss is about. After the encounter, consumers often dislike the artwork to some degree, or remain indifferent to it. Because people must buy the work before they experience it, there can't be a simple correlation between mass sales and mass mood. You and lots of others may be suckered into going to a film you dislike, but just by going you've already been counted as among those who support it. Doubtless many people enjoyed Spiderman. But it's very difficult to say why, at least if we want to move beyond claims about strategies of storytelling and cinematic presentation. And did all of the patrons enjoy it for the same reasons? That remains to be shown, and it's hard. We know that a movie may appeal to several audiences at once, packaging a range of appeals. Must we find reflections of cultural needs in every aspect of a movie that might appeal to someone?
A primary explanatory prop for reflectionists is politics. Talk about an American film of the 1950s, and sooner or later you'll invoke the reign of blandness that was the
31
Eisenhower administration. But why do we assume that America's mind-set switches its course whenever a new president is elected? What percentage of the electorate votes? And what percentage goes to movies regularly? And do these demographics overlap? It's well-known that a large slice of the audience since the 1960s consists of people too young to vote. So how are anyone's anxieties about presidential policies being reflected in the works the kids consume?
In addition, the reflectionist typically ignores the range of incompatible material on offer. If 1940s film noir reflects some angst in the American psyche, how to explain the audience's embrace of sunny MGM musicals and lightweight comedies in exactly the same years, indeed on the same double-feature program as a murky noir? The year 1956 saw the release of The Ten Commandments, Around the World in 80 Days, Giant, The King and J, Guys and Dolls, Picnic, War and Peace, Moby Dick, and The Searchers. Pick any one, find some thematic concerns there that resonate with contemporaneous social life, and you have a case for any state you wish to ascribe to the collective psyche. But take any other film, or indeed the industry's entire output, and you have a problem. The alternatives are to find common themes of an insipid generality or to float the rather uncompelling claim that several hundred films reflect many different, and contradictory, facets of the audience's inner life.
Moreover, reflectionists have always been reluctant to offer a concrete causal account of how widely held attitudes or anxieties within an audience could find their way into artworks. This is one reason that the usual invocation of presidential terms is unsatisfactory. Through what specific causal processes could changing the occupant of the White House affect popular culture? How exactly does a party platform or a candidate's charisma get translated into Hollywood movies for the multitudes? Furthermore, if there ever were a dominant mood at large in the land, it would be very difficult for that mood to be expressed in a current movie. There's often a lag of several years before a script finds its way to the screen; many of the films released in 1997, though read as responding to current crises, were bought as projects in 1993 and 1994. More important, movies are made by particular people, all with varying agendas, and they are inevitably going to shape the initial project in particular ways. Thus the preoccupations of the screenwriter, the producer, the director, and the stars rework the given idea. And these workers, we are constantly reminded, are far from typical, living their superficial lives in Beverly Hills. How can the fears and yearnings of the masses be adequately "reflected" once these atypical individuals have finished with the product? It now seems likely, for instance, that the violence in American films "reflects" not the taste of the mass audience but the egos of the makers, who enjoy the bravado of seeming to push the envelope.
In sum, reflectionist criticism throws out loose and intuitive connections between film and society without offering concrete explanations that can be argued explicitly. It relies on spurious and far-fetched correlations between films and social or political events. It neglects damaging counterexamples. It assumes that popular culture is the audience talking to itself, without interference or distortion from the makers and the social institutions they inhabit. And the causal forces invoked—a spirit of the time,
32
Poetics of Cinema
a national mood, and collective anxieties—may exist only as reified abstractions that the commentator turns into historical agents.52
It comes down, very often, to realizing that large-scale events need not have large-scale and distant causes, and small and medium-size events can have small and proximate causes. After 1920, the deep-space tableau style on display in films like Ingeborg Holm (Figures 1.1-1.2) went extinct, and an American-style continuity cinema came to the dominance it still enjoys. There's no need on the face of it to posit worldwide social upheavals as the direct cause for the new style. Some large-scale events, such as World War I's damaging effects on European film industries, surely served as preconditions, but those didn't directly cause the aesthetic changes. More proximate causes included the renewal of national film industries, the saliency of new templates of cinematic storytelling (such as the feature-length fictional narrative), and the emergence of a younger generation of filmmakers attuned to what seemed cutting-edge technique. These and other more proximate factors go a long way toward explaining the worldwide absorption of continuity premises. Likewise, the style has remained constant in its essentials for about a century, in the face of profound social, political, and economic changes—largely, we suppose, because it continues to fulfill functions that filmmakers deem worthwhile.
This isn't to say that society has no impact on films. Of course it does. But that impact isn't single or simple.53 I'm proposing that causal explanation in poetics can best proceed in steady steps, moving from the artwork to the proximate conditions of production (agents, institutions, and communal norms and practices). These in turn may be influenced by both immediate social causes and longer-term preconditions; we have to look and see exactly how. That Japanese films of the Pacific War period were shaped by political demands is undeniable, but the works of Ozu, Mizoguchi, and their peers reworked the assigned materials in distinctive ways. Likewise, the vague demands of socialist realism in the USSR were fulfilled in ways that both reject and rework Soviet montage norms of the 1920s. In any instance, a social command will be mediated by the film industry, existing traditions, and the varied ingenuity of filmmakers. Similarly, long-standing social attitudes, such as racism or homophobia, supply stereotypes, but those can be transformed by the process of production and the dynamics of the particular film. Poetics is in a good position to show how that works. It reminds us that themes will be recast, form and style will transform social givens, and filmmakers will still choose among importantly different ways to tell the story. Ideology doesn't switch on the camera.
From Shriek to Shot
I've suggested that all the constructional factors are connected, that inquiring into particulars and patterns can lead us to principles, purposes, and practices. To illustrate this, let me provide a tentative example. Before the arrival of videotape, Alfred Hitchcock's Rope (1948) was a mythical beast, unavailable in 16mm and seldom shown on television. On a trip to a film archive in the mid-1970s, I watched it, eager to enjoy what had become celebrated as one of Hitchcock's greatest experiments. The film tells
Figure 1.4 Rope (1948): The first "invisible" cut from Brandon's back...
Figure 1.5 ... disguises a change in the camera reels, which will be spliced together for projection on a bigger reel.
how two gay college men, Brandon and Phillip, garrote their friend David and hide his body in a decorative chest. They then invite his friends and relations to a dinner party at which the buffet is arrayed, ghoulishly enough, on the chest. The killers' former prep school teacher Rupert eventually realizes their crime and reveals it to the police.
What made everyone keen to see Rope was its reputation as a technical tour de force. Before production Hitchcock had announced that each shot would last nearly a full camera reel. A camera reel normally held 1,000 feet of film, but the Technicolor camera's maximum capacity was 952 feet, and some of that had to be wasted getting the camera up to speed, so the maximum length of a take for Rope would be around 10 minutes. The publicity for the film had made it seem that every shot was about the same length. "Each take averaged 925 feet," proclaimed an article in American Cinematographer, which also mentioned that the scenes were rehearsed in sections of about 9 minutes each.54 For decades thereafter, the 80-minute film was described as ' consisting of eight 10-minute takes.
Things are more complicated than that, because the film contains 11 shots, and their lengths vary considerably. Just as important are the ways in which the shots are connected. Rope begins with a comparatively brief high-angle shot establishing the street outside, seen under the credits, before the camera pans to a window and we hear a cry. The next shot is a very long take, but the shots that follow aren't all the same length, and most run significantly shorter than 10 minutes. More striking is the fact that some cuts are disguised by stretches of blackness, typically when the camera moves up to a character's back and away again. Yet other cuts are conventional eyeline matches: when a character looks offscreen, we get a cut to another character.
Is there a pattern to these particulars? After the external establishing shot, there's a cut inside to the parlor, where Brandon and Phillip are strangling David. This shot plays out for several minutes while the two young men jam their victim into the chest, close it, and start to talk. At the shot's end, we get a "hidden cut" as Brandon's back blocks the frame and a new shot starts (Figures 1.4-1.5). But at the end of that shot, number 3, there's a conventional eyeline-match cut. Kenneth looks off left—cut—a
34
Poetics of Cinema
Figure 1.6 Rope: The first visible cut, from Ken...
Figure 1.7   ... to the two killers leaving.
shot shows the two killers walking to the door (Figures 1.6-1.7). The rest of the movie follows this pattern. A long take ends with a blackout cut, then the following take ends with a visible cut. To get schematic about it, with slashes indicating the visible cuts:
Shot 1 (the exterior)/ Shot 2—blackout cut—Shot 3/ Shot 4—blackout cut—Shot 5/ Shot 6—blackout cut—Shot 7/ Shot 8—blackout cut—Shot 9/ Shot 10—blackout cut—Shot 11
But this pattern raises a question. Why disguise some cuts and not others? A viewer might mistake the shots linked by blackouts as all one take, but there was no effort to hide their neighbors (3/4, 5/6,7/8, and 9/10). Why not black out all the cuts?
The answer lies in exhibition practices.55 Today a film is usually mounted on a single big platter and run through one projector. But before the 1980s, theaters used two projectors, with the projectionist switching between machines to project one reel after another. Although camera reels held at most 1,000 feet of film, projection reels held 2000 feet (a maximum of 22 minutes running time). Knowing that the film would be projected on five double reels, Hitchcock shrewdly created the blackout cuts (linking shots 2-3,4-5,6-7, 8-9, and 10-11) for shots that would be spliced together on a reel. But there would be no similar way to disguise the reel change from one projector to another. Then, too, the start of each 2,000-foot projection reel would suffer some wear and tear, so hidden cuts wouldn't survive repeated projections. Hitchcock reconciled himself to presenting visible cuts between the shots that would be run on different projectors.56
Like many artists, Hitchcock submitted himself to fairly strict constraints to see what he could make of them. He created, we might say, fresh problems in order to find idiosyncratic solutions. But production practices and projection technology governed his choices only in an external sense. That is, nothing about camera magazines and projector reel changes dictated his finer-grained decisions about precisely how long
Figure 1.8 Rope: Phillip's wounded reply that he wasn't afraid to strangle a chicken yields a cut to ...
Figure 1.9 ... Rupert watching thoughtfully. It's the beginning of his suspicions, and the cut marks a shift in point of view, attaching us to him during the next phase of the film.
a shot should last or what should be shown at the end of one shot and the beginning of another. A detailed analysis of this remarkable film would take me too far afield here, but we might notice that Hitchcock clearly didn't use the "10-minute take" as an invariable yardstick. Even putting aside the establishing shot under the credits (a little over 2 minutes), the long takes vary in length quite a bit. Three shots run approximately 10 minutes, five last between 7.15 and 8.11 minutes, the introductory shot runs a little over 2 minutes, and the last two shots are comparatively brief, running 4.6 minutes and 5.6 minutes.57 Miklos Jancso, in Sirocco (1969) and Electra (1974), sustained each shot until the camera reel nearly ran out, but Hitchcock timed his cuts to articulate the unfolding drama.38
The most evident instances are the eyeline-match cuts. Hitchcock gives this normalized device a fresh force, not only because any cuts at all are rare but also because these create a powerful progression. The first shot change is low-key, when the guest Ken is left bewildered by Brandon's casual suggestion that tonight Ken might rekindle a romantic spark in his former girlfriend Janet. Ken watches Brandon and Phillip walk away assuredly (Figures 1.6-1.7). The next eyeline-match cut, between shots 5 and 6, is more dramatically charged. Phillip has just blurted out that Brandon s account of him killing chickens is a lie, and the cut takes us to their teacher Rupert, watching appraisingly (Figures 1.8-1.9). Hitchcock puts Brandon and Phillip's ensuing quarrel offscreen as we are allowed to study Rupert's reaction, a mixture of bemusement and wariness. This cut launches the central portion of the film (and the third projection reel), when Rupert's suspicions steadily grow. During this reel and the next, the attached point of view shifts from the murderous couple to Rupert, who scrutinizes them and questions the maid, Mrs. Wilson. The next eyeline cut occurs when Rupert and his two pupils look offscreen in reply to the maid's announcement of a phone call. Just before the cut, Rupert is telling the killers that there's something
36
Poetics of Cinema
Poetics of Cinema
37
Figure 1.10 Rope: Phillip's hand clutching the revolver ...
Figure 1.11 ... marks another high point, as Rupert realizes that his pupils are capable of violence.
going on that's upsetting them a great deal. The cut marks another stage in Rupert's growing sense that they've done something reprehensible.
The last visible cut forms a kind of climax. Rupert has just been imagining how the murder might have been enacted, with the camera tracing the path of the action, as if following his gaze. The framing comes to rest on Brandon's pocket, where his hand clutches a pistol (Figure 1.10). Cut to Rupert, staring (Figure 1.11). Across the film, the shot joins have set up an internal norm, with the eyeline match presumed to be the prime linkage device. All these cuts show symmetrical variation too. The first and third cuts are motivated by a glance at the end of a shot. In the second and fourth cuts, the object of a glance—a distraught Phillip, a pistol in a pocket—ends the shot, and the follow-up starts with a character or characters looking off. In the first cases, the cut is perhaps somewhat more predictable, because a close view of a character's look sets up the expectation of an eyeline match. In the other instances, the cut becomes more unexpected and interruptive. After all, either the close-up of Phillip or that of the pistol could easily be part of a sustained shot panning among the characters.
A parallel progression is provided by the blackout cuts. The first one (shots 2-3) and the third one (shots 6-7) are motivated by tracking in and panning past Brandon's back, with his jacket blotting out the cut (Figures 1.4-1.5). The second blackout executes the same maneuver, this time using Ken's back (shots 4-5). Again, an internal norm is set up. Even the viewer who isn't keeping strict track will probably come to expect that these somewhat contrived blackouts will be part of the film's stylistic unfolding. Unlike the eyeline matches, the first two cuts don't mark distinct phases of the drama; they all continue scenic action fairly fluently. The third blackout cut is more marked, occurring as Rupert mentions David's absence to Brandon. Over the black frames we hear Rupert say, "As a matter of fact, I'm beginning to miss him myself."59 The final blackout cut gathers the most force. When Rupert realizes that David's body may be in the chest, he pulls up the lid and the camera tracks abruptly forward. As the chest fills the screen, the image goes dark. Coming out of the transition, the camera lifts over the edge of the lid to reveal Rupert's sickened expression
Figure 1.12 Rope: Rupert lifts the lid and discovers David's body in the chest.
(Figure 1.12). At this climax the blackout gains visceral impact, suggesting what can't be shown and providing a purely graphic thrust from light to darkness to light, as Rupert's stricken face fills the frame. The shots linked by this ominous blackout are the film's shortest ones, accelerating the film's denouement.
We've identified some localized patterns and functions, but what's the broader purpose? Why did Hitchcock go to all this trouble to sustain lengthy takes? Why try to make a film with so few cuts? The choice is particularly odd in that Hitchcock had long taken pride in his mastery of editing. He famously proclaimed, "If I have to shoot a long scene continuously I always feel I am losing grip on it, from a cinematic point of view.... What I like to do always is to photograph just the little bits of a scene that I really need for building up a sequence."60 Until Rope, Hitchcock indulged in a flashy long take now and then (Young and Innocent, Notorious), but across a whole film his cutting rate tended to be fairly fast. Most Hollywood features of the 1940s had an average shot length (ASL) of 8-11 seconds. Sometimes Hitchcock's work fell in that range, but he was also inclined toward a more brisk decoupage.61 Foreign Correspondent (1940), Saboteur (1942), Lifeboat (1944), and Notorious (1946) all have ASLs falling between 6 and 7 seconds. Hitchcock's producer, David O. Selznick, thought that his films tended,to be "cutty" and sometimes tried to slow the editing pace by replacing Hitchcock's single close-ups with more sedate two shots.62 In 1947, The Paradine Case (1947), which Selznick recut, averaged 7.3 seconds per shot. The following year, Rope's shots averaged 7.3 minutes. So again, why did Hitchcock change his style so radically?
Probably several factors worked together. Hitchcock claimed that the long-take procedure saved money, not an unimportant element at a moment when he was considered a somewhat budget-straining director. On Rope, his first venture as an independent producer, he may have wanted to show that he kept an eye on the bottom line. Still, throughout his career he had presented himself as fiscally efficient, largely because he claimed to plan each film out on paper to the smallest detail. Press accounts of the tension-filled Rope set, in which any missed cue or bungled line forced the whole production to start the shot over, suggest a precarious undertaking that a
38
Poetics of Cinema
Poetics of Cinema
39
prudent man wouldn't try. In any event, from the standpoint of historical poetics we can propose that another factor was at work.
New norms were emerging in the 1940s, and some directors were making flamboyant use of them. Welles' Citizen Kane (1941) had treated long takes not just as stylistic flourishes but also as dramaturgical building blocks. A scene might consist of a string of long takes, or one long take alongside a few briefer shots. On rare occasions the scene consisted of just one long take, what the French came to call the plan-sequence or single-shot sequence. In Kane's follow-up, The Magnificent Ambersons (1942), a late-night kitchen conversation consumes a single shot lasting 5 minutes. Once the self-conscious long take was on the agenda, some filmmakers tried to test the resources of the device. Otto Preminger, George Cukor, Vincente Minnelli, and others made the long take a cornerstone of their filmmaking. Joseph Mankiewicz's The Late George Apley (1947) has an ASL of 16 seconds, whereas Preminger's Fallen Angel (1945) clocks in at 33 seconds. The most flagrant experiment along these lines was the pseudo-subjective movie The Lady in the Lake (1947), which averages over 2 minutes per shot. As if not to be outdone by such upstarts, Welles' Gothic Macbeth (1948) presents the murder of King Duncan in a single camera reel running nearly 11 minutes.63
Most of the long-take films of the period also made use of the mobile camera. Again, Welles' flamboyant style made the option prominent, notably in the ballroom scenes of Ambersons. A few years later, intricate tracking shots were made easier thanks to several new cranes and dollies, notably the crab dolly; its tight turning radius allowed the camera to spiral around a prop or an actor. Observers inside Hollywood's technical community began noticing the "fluid camera" technique. Joseph LaShelle, who shot Laura and Fallen Angel for Preminger, was praised for
breaking a scene down into various forceful compositions and joining these different 'points of view' together through smooth camera movement— On the screen a close-up gives way to a long shot which then evolves into a follow shot.64
This is exactly what some observers would claim that Rope does—translate orthodox editing patterns into panning and tracking movements that connect distinct camera setups.
We tend to think of Hitchcock as a self-motivated innovator, but he seems to have been highly sensitive to what his peers were up to. He sometimes adopted the aggressive foregrounds and deep staging popularized by Kane (Figure 1.13). Likewise, Selznick encouraged him to shoot longer takes, and it seems likely that the emerging "fluid camera" aesthetic aroused the competitive instincts of Hollywood's most self-conscious experimentalist. The film Hitchcock directed just before Rope was The Paradine Case (1947), for which Morris Rosen, the head grip, devised an early crab dolly. (Rosen would operate the camera boom on Rope.) The Paradine Case included several lengthy shots, and one was trumpeted in a technical journal as the "Three and a Half Minute Take" (though it didn't survive Selznick's final cut).65 Hitchcock's rationale, as paraphrased by one reporter, amounted to a repudiation of his cutting-based aesthetic. "The big advantage gained artistically is the simulation of stage continuity.... Too often in the past, [Hitchcock] believes, a good dramatic picture is hampered by too frequent cuts, not
Figure 1.13   In Notorious (1946), Hitchcock    Figure 1.14   Striking frontality and compact works a variant on the aggressive foregrounds    arrangement in Lifeboat (1944). of Wyler and Welles.
enough continuous action."66 Hitchcock likewise defended Rope's style on the grounds that it preserved a stage play's fluidity, but he acknowledged that the technical challenge was another attraction for him. "Right from the beginning," recalled screenwriter Arthur Laurents, "he'd tell me that he was going to do it as a play, and with, I think, nine takes, or nine reels—something like that. And that interested him because that hadn't been done."67 Hitchcock, in effect, created a new set of problems for himself. He explained to Francois Truffaut, "I undertook Rope as a stunt."68
The stunt blends other Hitchcock preoccupations. In previous American films he experimented with what we might call the floating close-up, a prolonged tracking shot attached to a player's face. This satisfied Selznick's concern for sustained takes and glamorous portraits of female stars,69 but Hitchcock tended to use them for suspense, : often locking them within editing patterns that indicated the character's moving point of view. He also began huddling his characters close to one another in a tight medium Shot, often turning all their faces to the camera in a quite artificial way (Figure 1.14). This staging strategy is very salient in Rope (Figures 1.15-1.16), largely because without cutting Hitchcock can't easily alternate over-the-shoulder reverse angles. Furthermore, Rope's sequences were shot in story order. Hitchcock had experimented with this tactic in The Paradine Case. He shot the film chronologically, and several of the courtroom exchanges were filmed in real time with four cameras running simultaneously.70 Similarly, Hitchcock had already explored the possibility of restricting a film's action to one setting—in a partial way during the train scenes of The Lady Vanishes (1938), and more systematically in Lifeboat (1944). Rope offered a chance to try the idea again. According to a contemporary report, "The idea [of a long-take film] had been one of Hitchcock's pet dreams for a long time. But he needed a story that had no time lapses, and a story that took place on one set."71 Patrick Hamilton's play, whose three acts aren't broken by time gaps, became the basis of a bravura synthesis of long takes and camera movement within a single locale, all presented in strict continuity.
True to Hollywood's alibi for formal experiment—the story is all—Hitchcock claimed at the time that the viewer shouldn't notice the outre" technique, "The
40
Figure 1.15 Rope: In the absence of over-the-shoulder staging, the characters huddle frontally.
Figure 1.17 Rope: At the climax, Rupert unfolds the rope while commenting that he won't ride with his pupils.
Poetics of Cinema
Figure 1.16 Rope: Characters strung out horizontally. The depth relations couldn't be as baroque as in Figure 1.13 because lighting for color film didn't permit the extreme focal distances available for black and white.
Figure 1.18 Rope: The searching camera scans the room as we hear Rupert speak of suspense...
audience must never be conscious of it.... The result I'm after is to excite the audience by making the picture flow smoother and faster."72 Yet the blackout cuts call attention to themselves, and several critics of 1948 pointed out that what Hitchcock called the "roving camera" could be quite obtrusive.73 Some moments flaunt the prolonged take quite explicitly. The camera's anxious probing of the apartment during Rupert's voice-over replay of the murder is one instance (another device Hitchcock had tried out earlier, when Maxim recounts his wife's death in Rebecca). An even more overt passage comes when Rupert fishes the rope out of his pocket, turns to his pupils, and extends it (Figure 1.17). Over a close-up of his hands, we hear his voice say, "Driving with you and Phillip now might have an additional element of... suspense." Before the last word has been spoken, the camera starts to glide rightward away from him, passing a corner and a window (Figure 1.18) before settling on the faces of Brandon and Phillip (Figure 1.19). The refusal to cut obliges the camera to traverse the space completely and make us wait for the men's response.
In an ordinary film, using a tracking shot to postpone their reactions would seem ham-fisted, but such self-initiated camera movements, independent of character
Poetics of Cinema 41
Figure 1.19 ... and settles on the two killers, in a long-take version of the eyeline matching that has linked shots through cuts earlier (Figures 1.6-1.11).
movement, have become prominent in the later phases of this film, so this tactic seems a logical culmination. Naming the shot's overt purpose—to provide "an additional element of suspense"—on the dialogue track marks the device even more explicitly. It might seem merely a critic's fancy that the lethal rope is a kind of emblem for the stretched-out take and the unwinding camera movements, but a poster advertising the film reinforces the link. The taglines play up the opening, the closing, and the sinuous continuity in between: "It begins with a shriek!... It ends with a shot! From beginning to end, nothing ever held you like Hitchcock's Rope."
Having made the longest-take movie in studio history, Hitchcock could afford to let up a bit. His follow-up, Under Capricorn (1949), averaged 44 seconds per shot (placing it second, I believe, in Hollywood's long-take sweepstakes). The most sustained shot, when the distraught Henrietta tells the protagonist of the secrets in her past, runs about 8.5 minutes. Although that sequence is relatively stationary, relying on subtle reframing and refocusing, for other shots Hitchcock posed himself new problems, tracking characters through several rooms and up and down staircases.74 Under Capricorn often uses editing to ratchet up the dramatic tension, as we saw with the eyeline matches in Rope, but it also revives the concealed-cut device, giving a subliminal sense that a shot runs even longer than it did on the set. Across film history, directors tend to love long takes, but producers hate them because scenes can't be tightened up in postproduction. With The Paradine Case, Hitchcock was at Selznick's mercy, but he produced both Rope and Under Capricorn himself and was able to indulge his experimental ambitions. Naturally, he made sure that the films' bravura techniques were fully covered in the press, both popular and professional.
It would be worth studying how the problem of filming theater was being rethought by several other directors at the time. Bazin wrote a brilliant essay about this development, suggesting that Laurence Olivier's Henry V (1944) and Jean Cocteau's Les Parents terribles (1948) had devised various cinematic equivalents for the experience of staged drama. Bazin noted that Olivier starts his film with a performance before an audience, whereas Cocteau confines his camera to an apartment in an
42
Poetics of Cinema
effort to suggest the suffocation of the play's single set,75 One might add Jean-Pierre Melville's Les Enfants terribles (1950), which flaunts theatrical interpolations like a descending curtain. Wyler's The Heiress (1949) and Detective Story (1951) offer less flamboyant but no less intriguing instances. Like Bazin, Wyler speaks of trying to steer a middle course between simply photographing a stage play and opening it up so broadly that it loses its theatrical flavor. In Detective Story, "I did not change the construction of the play; I changed the set. Instead of having two little rooms, as in the play, I have five or six rooms."76 Hitchcock made an effort along similar lines with Dial M for Murder (1954). Because "the film will have to follow the play very closely," he explained, "I am treating it in a modified Rope style."77 DialM boasts an ASL of 9.1 seconds (the same as Rebecca and Spellbound) and trim, efficient staging. Hitchcock obeys the Bazin-Wyler dictum, opening up the play only a little by showing the apartment's terrace, bedroom, and outside hallway.78 The absolute confinement of Rope had defined one pole of theatrical cinema, and most directors, including Hitchcock himself, retreated to a middle way.
As for the long take, Hitchcock and most of his peers abandoned it as a structural unit in the 1950s. Yet the dream had great staying power. The neorealist screenwriter Cesare Zavattini imagined 90 minutes of real life presented in a single shot.79 Many people believe, mistakenly, that Andy Warhol films like Sleep (1963) and Empire (1964) consist of static takes many hours long. Aleksandr Sokurov's Russian Ark (2002) uses digital video (and an orchestration of action at least as intricate as that in Rope) to capture uninterrupted time. Josh Becker's low-budget feature Running Time (1997) resorts to hiding its cuts in the Hitchcock manner. In the 1940s, filmmakers began to acknowledge shot duration as a formal parameter, and we might conjecture that Hitchcock, like many of his peers, took the long take as a challenge, an occasion to reshape contemporary norms of cinematic storytelling.
I could have introduced this brief analysis of Rope's decoupage by identifying the general problem Hitchcock set himself, that of a feature film presented in something approximating a continuous shot. I could then have discussed the functional consequences of this goal, including the new problems it posed (the absence of reverse angles, the need to allow reel changes) and the rational solutions that Hitchcock found (frontal staging, hidden cuts). Instead, just to give the flavor of an inquiry into poetics, I tried to show how the analyst can frame and revise questions that move among several pertinent aspects of a film, from details to patterns, from functions to principles, and from internal dynamics to historical context.
I wanted as well to highlight how the rubric of practices includes the institutional forces at work, like production and projection routines, and the technology employed, such as standardized reel lengths. A further lesson here is that practices include informal relations among personnel, such as the urge to show one's skill. Filmmakers may address their work to their peers as well as to their public. No historian of painting would be surprised to learn that artists compete in displaying their virtuosity. We can explain important aspects of how movies work by considering filmmakers as creative agents working with craft practices within a community. Members of that community may be sharply aware of traditions and trends. They may replicate well-tried
n
Poetics of Cinema
43
norms or explore emerging ones. They may solve problems in routine ways or pose new difficulties in order to triumph over them. And some of the most ambitious and gifted creators are likely to treat constraints as opportunities.
What Snakes, Eagles, and RKesus Macaques Can TeacH Us
I once projected a kung fu film, Snake in the Eagle's Shadow (Yuen Woo-ping, 1978), in a Cantonese-language print lacking both English dubbing and subtitles. The question was, How much of the film could the audience grasp without knowing its native language?
Actually, quite a lot. It might seem too obvious to mention, but we in the audience perceived the film. In Snake in the Eagle's Shadow, we saw patches of color, patterns of light and dark, trajectories of movement, and changing shot displays. But our perception wasn't really of abstract configurations. Humans evolved to detect objects and actions in a three-dimensional world, and in watching Snake we definitely recognized things. We saw young and old men and women, all going about activities in a voluminous space. We saw a youth involved in social interactions, in locales—a village, a clearing—that we could recognize, at least generically. We heard speech, and if it had been in English, we could have grasped it as quickly and involuntarily as we grasped the sight of a human face. We heard noises, such as the blow of fists on flesh, or labored breathing, or the sound of a cobra hissing. We heard music, mostly in a tradition we recognized, and it registered as such.
But also, and more interestingly, we viewers of Snake in the Eagle's Shadow understood a lot of the story. We understood that the protagonist was a young servant in a kung fu school. He wants to learn martial arts and meets an old man who, despite his shabby appearance, is a master fighter. The youth undergoes arduous training and eventually comes to defeat a villainous master. These features aren't simply given in perception; we had to bring in large domains of knowledge to arrive at this story. Viewers who were familiar with the kung fu genre could structure the film along familiar lines, but even those who weren't martial arts fans understood a good deal of the action because they had skills in understanding any type of story. At one point, when the protagonist sees a cat fight a cobra, all of us realized that he was inspired to model his kung fu technique on the cat's attack. Call our activity of this sort comprehension, a grasp of the concrete significance of the perceptual material as patterns of social action. In this case, the patterns are presented in the form of a story.
Finally, spectators used the film in various ways. Snake in the Eagle's Shadow wasn't intended to be shown in a college classroom, but I drew it into my own agenda. Some of the students took the film as an occasion to celebrate the prowess of Jackie Chan. Others took it as proof of the artistic bankruptcy of Hong Kong cinema. Some students from Hong Kong read it as a statement of local pride in the face of adversity. Those who practiced martial arts themselves spotted techniques that they could try themselves. Let's say that all of us appropriated the film, in however disparate ways.
44
Poetics of Cinema
These types of activity suggest how poetics can address what I'm calling the processing of films by viewers. If poetics is concerned with how filmmakers use the film medium to achieve effects on spectators, we ought to have some idea of how those effects might be registered. Film researchers aren't psychologists or sociologists, but we can draw upon the best scientific findings we have to mount a plausible framework for considering effects. The poetics I propose is thus mentalistic. It assumes that we can characterize the spectator's embodied mind as engaging with the film. It's also naturalistic, pxesuming.that scientific investigation of mental life is likely to deliver the most reliable knowledge. I'd also propose that the best mentalistic and naturalistic framework we have available is that provided by what we can broadly call the cognitive approach to mental life.
Adopting this perspective makes some researchers worried. Some object that it neglects the influence of society, ideology, or culture on viewers. But this is to assume that a mentalistic and naturalistic framework focuses wholly on individuals. It doesn't. Cultural activities are mental in an important sense: They're learned, recalled, rethought, and so on by the embodied minds of social agents. The framework presumes some intersubjective regularities of mental activity across individuals, but cultural theorists do the same thing when they discuss how members of a subculture come up with a resisting reading. Other critics have argued that conceiving of the spectator in the way I propose neglects the differences of race, gender, ethnicity, and other markers of identity. Yet clearly there are common effects across such groups; people of all sorts feel suspense in a thriller and sadness in a melodrama. Studying such commonalities isn't on the face of it unreasonable or uninteresting. Moreover, there's always a degree of idealization in discussing spectatorship. Just as linguists create the idealization of "the native speaker" in order to understand grammatical principles, virtually all researchers are obliged to idealize the spectator, even the female or African American spectator, to some degree. Finally, although not every conceptual framework fits well with every cluster of research questions we might want to float, I think that some identity differences can be understood from the standpoint of poetics, as I'll try to show shortly.
We can start to understand the effects of films by borrowing a distinction from classic cognitive psychology, that of top-down and bottom-up mental processing. Top-down processing is concept driven; bottom-up processing is data driven. A classic instance of top-down processing is problem solving. Given a crossword puzzle, you draw upon your stored knowledge about language and the world (including the stratagems of crossword puzzle designers) to fill in the blanks properly. By contrast, bottom-up processing arises from a moment-by-moment encounter with the world. As you enter an unfamiliar room, for instance, your visual system picks up information about edges, brightness differences, and a host of other features that coalesce into a spatial whole.
Our brains can process information in both "directions" at the same time, so any particular experience will be a mixture. While searching your memory for the right word, you get bottom-up information about the crossword puzzle from the written clues and the array of empty spaces and black ziggurats on the page. Upon entering an
Poetics of Cinema
45
unfamiliar room you quickly rework the perceptual input in the light of knowledge, identifying a chair in the shape of a beanbag and a lamp bubbling like lava leads you to make a higher-level inference about the tastes of the people living there. On the whole, bottom-up processes are fast, involuntary, cheap in cognitive resources, and fairly consistent across observers.80 In an important sense, all TV viewers watching the horrendous crash of the airliners into the World Trade Center saw and heard the event in the same way. Top-down processes are slower, more voluntary, more expensive in cognitive resources, and more variable across observers. Having seen the Trade Towers assault, viewers interpreted its significance in different ways—as an act of war, as a response to globalization, and/or as a counterthrust to U.S. imperial ambitions.
Perceptual uptake occurs in milliseconds, and for good reasons. Our brains evolved in situations in which survival demanded reasonably accurate information about spatial layout and other agents. Consequently, the activity of our perceptual mechanisms is hidden from us; we can't watch our retinal image or our neuronal firings. And although experimental films like James Benning's Ten Skies (2004) create noticeable visual effects, like illusory movement, we can't really probe the mental hardware yielding the experience. Nearly as fast are intuitive judgments, as when we sense that a person is arrogant or kindly, or when we just know we'll like a class after hearing just a little of the teacher's opening lecture. Although we can make these judgments in a few seconds, they draw on stored knowledge and are thus to some degree top-down.81 Yet even these remain fairly impervious to introspection.
The top-down-bottom-up distinction drastically simplifies a complex process that would probably be best modeled along several dimensions rather than a single vertical one. Doubtless neurological research will eventually show that any experiential process involves complicated feedback and input-output among many mental
: systems. Take mirror neurons, which can be found in various areas of the brain. Watching someone lift a heavy weight, either in front of you or on a movie screen,
"stimulates some of the neurons in your brain that would fire if you lifted a weight yourself. Many of these mirror neurons are linked to intentional action on your part, so that when they fire, you can spontaneously understand the actions of others as products of their intentions. It seems that we have a powerful, dedicated system moving swiftly from the perception of action to empathetic mind-reading.82
This is only one instance of how contemporary research asks us to consider that many of what we take to be learned or culturally guided mental activities will turn out to be packed into our biological equipment. Psychological research in the cognitive paradigm has steadily diminished claims for a blank-slate conception of the human mind and belief in the unlimited plasticity of human capacities. More and more activities (e.g., language, recognition of emotional signals, and attribution of intentions) seem traceable to humans' supersensitive natural endowment. Many specialized faculties need only triggers from the regularities of our world to lock in and function at high levels quite quickly. As research goes on, many "higher-order" activities will probably be revealed as grounded in a rich perceptual system present at birth but awaiting activation and tuning from the environment.83
46
Poetics of Cinema
A Model of Viewer Activities
Concept-Driven processes
4 4 14
Appropriation
Comprehension
Perception
Phenomena! film
,jf > f
tit. tit
Í    t    t t
Data-Driven processes
Figure 1.20 A schematic model of the spectator's activities. Continuous arrows indicate the primary direction of processing, with dotted arrows indicating a degree of feedback among processes.
So I grant the schematic quality of my distinctions. As a first approximation, however, they can clarify how our minds interact with movies. I suggest that we can characterize viewers' interactions with films along a continuum of activities: perception, comprehension, and appropriation.84 (See Figure 1.20.) Sensory input drives perceptual processing; perceptual processing feeds into comprehension and appropriation, in the "bottom-up" direction. Appropriation drives comprehension to some degree and perception to a lesser degree. There are secondary feedback effects, too (indicated by the dotted arrows in Figure 1.20), as when the manner of appropriation can recast perception or comprehension. For example, a decision to interpret the film a certain way can lead us to look more closely at the film and notice or comprehend aspects that might otherwise be missed. I'd argue that such feedback systems can't go all the way down or all the way up, because perception can't in every respect determine appropriation, and appropriation can't completely reshape perception or comprehension. Wishing that Thelma and Louise don't die won't make it so.
Perception evolved in large part to give us reliable information about the three-dimensional world in which we live. Representational films solicit this activity straightforwardly: We involuntarily see the world depicted on the screen. We recognize our conspecifks and their surroundings. We hear noises, music, and language. We also see movement where there is only a stream of rapidly projected still pictures.
1
- -
5.
Poetics of Cinema
47
The perception of film as a representation of the world emerges very early in human development. Many of the experiments testing babies* reactions to facial expressions use televised images of the mother, which indicates that children spontaneously identify an audiovisual representation of the caregiver.85 Paul Messaris has shown convincingly that people in cultures without images recognize films and photographs as presenting persons, places, and things.86 The perceptual mechanisms that film engages seem to be shared with other primates. Experimenters routinely use videos to test chimps, monkeys, and their cousins, and the results indicate that these creatures identify their counterparts on the screen. Other experiments suggest that perception is also attuned to displays of emotion. When rhesus macaques who are unafraid of snakes watch a film of other monkeys shrinking from a snake, they begin to show fear themselves.87
This isn't to say that the processes of filmic perception are innate, as if a newborn could enjoy Snake in the Eagle's Shadow. Perceptual development unfolds in response to the environment. Humans and other primates have evolved to be ready for first encounters with the world's regularities. By the time people watch movies with any de gree of perceptual understanding, they have developed the capacities to negotiate the three-dimensional world as well. And perception isn't terribly plastic. Given a stable input, perceptual capacities will develop along well-marked pathways. No one learns to see in infrared, or to move as if the world were two-dimensional, or to differentiate the separate frames of film flickering past. We are not so made.
The activities of filmic perception tend to be neglected by scholars today, but there's a long tradition of film aesthetics that places importance on the moment-by-moment effects of composition, lighting, cutting, and the like. From the Russian montagists through Rudolf Arnheim, André Bazin, and Noel Burch, theorists have paid attention to fine-grained creative choices that structure the viewer's perceptual uptake. Although strict experiments on filmic perception are welcome, there's a lot to be said by those of us not wearing lab coats. We can be sensitive to how patterns
.....andpractices of the medium shape such apparently simple strategies as directing the
viewer's attention. Some of the essays that follow try to show ways in which style shapes our moment-by-moment perception of the flow of images.
In comprehending a film, we construe the outputs of filmic perception as representing a hierarchical pattern of actions, a conception, or simply a train of sensuous elements (as in an abstract film). The viewer applies a wide range of knowledge to make sense of film, segment by segment or as a whole, and to give it some literal meaning. Narrative comprehension is the clearest instance. In my Snake experiment, spectators were able to build out of a perceived world a story about an ambitious young man who wants to master kung fu. Comprehension also comes into play when we're asked to grasp a cinematic argument or lyric. Comprehension is evidently a matter of degree; some viewers get more, some get less. In Snake, the viewers who knew Jackie Chan recognized him as the star and probably hypothesized that he would triumph through vigorous abuse of his body. Those who didn't recognize Jackie were probably surprised by the punishment he inflicted on himself. But the fact that comprehension varies in degree only indicates the extent to which it's a top-down process. Not everyone has the same set of conceptual schemes.
48
Poetics of Cinema
Again, poetics has a lot to contribute to understanding comprehension. The technical choices made by filmmakers organize perception in ways designed to enhance comprehension. Filmmakers design their shots and scenes so that spectators can follow the movie's large-scale form. Focusing on certain traditions or particular films, we can study how principles of style, narrative, and the like aim to provide a distinct experience for the viewer. This will be a central concern of most of the essays that follow. For instance, since, the 1920s most films in most countries have organized their perceptual surface according to some basic principles. Commercial storytelling cinema has long followed the conventions of analytical editing: master shot, followed by a two shot or over-the-shoulder shots, followed by singles highlighting each participant in shot/reverse-shot fashion. In fact, seldom do we find in any art a style with such pervasive presence and 100-year longevity. These norms have provided easy, comprehensible ways for narrative action to be understood.
Beyond stylistic patterning, it seems clear that comprehension also rides upon action schemas that the spectator can activate. When one man slaps another, and the second responds with a punch, there's not much doubt about what is going on in the story world at this point: insult and physical conflict. Likewise with a theft, an abduction, or glances that suggest attraction between man and woman. The large-scale form of the film is designed to create a flow of cues that ask viewers to apply schemas for typical situations and human actions and reactions, locking them into place quickly. Indeed, there's good reason to believe that these action schemas enable us to learn the stylistic schemas that present them. We know that people tend to face one another when they converse, so this regularity of social interaction makes comprehensible the stylistic option of shot/reverse-shot editing. As Gombrich puts it, "It is the meaning which leads us to the convention and not the convention which leads us to the meaning."88
The perceptual surface can also be so roughened that holistic action patterns become difficult to grasp. In The Two Minutes to Zero Trilogy (2003-4), Lewis Klahr presents shaky fragments drawn from comic books, showing only bits of words and imagery of bank holdups and police chases. The images are so broken up and they shudder past us so quickly that we never get time to figure out character relations or construct a complete story (Figure 1.21). The slender cues summon up action-based schemas but also frustrate our efforts to absorb them into scenes and larger narrative patterns.
Comprehension occupies an intermediary place in my framework, balancing between data-driven and concept-driven features. Appropriation is much more top-down. Here the viewer uses the film in a more or less deliberate way, drawing it into her personal projects, and she may stray far from the phenomenal film. I appropriated Snake in the Eagle's Shadow as a classroom example, but fanboys have appropriated it as a cult object. Films are appropriated by individuals and communities for all manner of purposes. People employ favorite films for mood management, watching Die Hard to pump themselves up or Sleepless in Seattle to have a good cry. Bloggers may use films to flaunt their tastes or strike a posture, whereas academics interpret films to validate a theory. Social groups appropriate films to a multitude of ends, treating some as praiseworthy representations of political positions and castigating
Poetics of Cinema
49
1
1 -
Figure 1.21 A bank robbery is evoked in jittery comic book imagery in The Two Minutes to Zero Trilogy (Lewis Klahr, 2004).
others as harmful. Filmmakers are always surprised by the range of ways in which people take their movies.
Accordingly, much of what interests cultural critics are acts of appropriation. Some Asian Americans have attacked Charlie Chan movies as exemplifying Hollywood racism, not only in their plots and characterizations but also because the Chinese Hawaiian detective is played by a Westerner in "yellowface." They have perceived the films and comprehended them; but they have appropriated them in a way very different than the makers intended, or could probably have imagined. Much of what Janet Staiger attributes to "perverse spectators" consists of unusual forms of appropriation. She writes,
Knowing that fictional narratives are produced permits many viewers to concentrate on narrational issues related to the production of the text. A study of ..some 1950s gay male viewers of A Star Is Born (1954) revealed that they were much more interested in constructing the story of the production of the film (when did Judy Garland shoot which scene) than in the film's plot—which at any rate was already "known."89
It's an interesting fact about films that groups (and individuals) can build unforeseen inferences out of particular aspects of a film that interest them. Nonetheless, what Staiger calls "uncooperative spectators" tend to perceive and comprehend the film in quite convergent ways, as she indicates by saying that the gay audience already grasped the film's plot.
Staiger's example typifies the tendency of cultural critics to stress divergence of response among groups. We know as well that there's likely to be considerable differences among individuals in any group we pick out. Both sorts of divergence are explicable in the light of top-down appropriation, for in this domain an indefinitely large range of conceptual schemes can be brought to bear on any phenomenon. A balanced account will also note the high amount of convergence at all levels. Academic interpretations display great agreement at the levels of perception and
50
Poetics of Cinema
comprehension, as well as a surprising degree of overlap in interpretations too, as when one critic revises the reading of her or his predecessors.90 More importantly, there are many convergences among spectators at the level of comprehension. All storytelling traditions evidently deploy such concepts as protagonist, goals, personal agency, conflict, and causal change—all concepts relevant to comprehension. Patrick Hogan has shown that some prototypical narrative patterns, such as romantic tragicomedy, are to be found throughout the world's literatures.91
This three-stage framework helps us understand the range of control available to filmmakers and viewers. Critics often ask, How much does "the text" control its "readings"? This framework lets us give some focused answers. As we move up the ladder, from bottom-up to top-down processing, the filmmaker's control diminishes and the spectator's power increases.
By constructing the phenomenal film, the filmmakers control very strongly, though not absolutely, the viewer's perception of it. It's impossible for a viewer to perceive the hero of Snake in the Eagle's Shadow as, say, tall and blonde. In addition, all the options of film style and structure can be mobilized to guide the viewer's notice to certain material. Framing can center key information, whereas cutting can highlight a detail. Style operates at all levels, but among its basic tasks is organizing the stimulus for uptake—even if that uptake is made difficult by an oblique technical choice or a problematic narrative.
At the level of comprehension, the filmmaker still has a lot of control, because features of theme and subject, style, and large-scale form are mobilized to guide the spectator's overarching understanding of the material. Still, no formal pattern can anticipate every question that can be asked about it. In grasping narrative form, for instance, the spectator contributes a lot—picking up the cues planted by the filmmakers, as well as inferring, extrapolating, filling in gaps, and the like. Most of this inferential elaboration is foreseen and governed by the filmmaker, but not all of it is. Shakespeare famously leaves us with the question of whether Macbeth and Lady Macbeth ever had children. She claims she has "given suck" to a babe, but Macduff says, apparently of Macbeth, "He has no children." Did the couple have a child who died? Did Lady Macbeth suckle another woman's baby? Spectators will differ in the ways that they deal with such zones of indeterminacy. Most will ignore them, but others will use them as occasions for appropriation, as when fans write fiction filling in gaps in Star Trek or The Lord of the Rings. No narrative can avoid leaving some openings for inferential elaboration of this sort. Louisa May Alcott couldn't have anticipated Geraldine Brooks' novel March, a fictional biography of the father depicted in Little Women. Of course, some artworks deliberately introduce gaps that they decline to fill, as with ambiguous endings.
Clearly, filmmakers have the least control over the activities I've gathered under the rubric of appropriation. Having perceived and comprehended (to a greater or lesser degree) Snake in the Eagle's Shadow, you're free to do with it as you will. Viewers sympathetic to gay rights can take Brokeback Mountain as a plea for tolerance, whereas those opposing gay rights can treat it as Hollywood propaganda for alternative lifestyles. Director Ang Lee and his colleagues can seek to shape the films
■J-
1
m
Poetics of Cinema
51
1
reception through publicity campaigns, but they can't anticipate every way the movie can be appropriated. The wave of mashup trailers that swept the Web (Brokeback to the Future, The Empire Brokeback) weren't foreseen by the filmmakers, however much they may have welcomed them. In sum, as we move up my chart, filmmakers' freedom wanes and spectators' power increases.
I haven't mentioned a prime component of film's effects: emotions. In my past work, the historical poetics I've proposed has slighted emotions, leading some people to think that a cognitive perspective can't tackle such matters. In part, leaving emotion out of the picture is a simple piecemeal idealization of the phenomenon; studying the grammar of a joke may not yield insights about what makes it funny. In addition, in the early 1980s, when the cognitive perspective was hitting critical mass in several
- disciplines and when I imported some of its observations into film studies, emotional matters were set to one side. But they weren't legislated out of existence, and the 1990s saw vigorous efforts to incorporate emotional life into the cognitive framework. This was also reflected in film studies, in the work of Murray Smith, Ed Tan, Torben Grodal, Greg Smith, Carl Plantinga, and many other researchers since.92 It's not an area of specialization for me, but I think that the processing framework I've proposed can accommodate emotion as an integral part of a film's effects.
: Emotion is part of our evolutionary heritage, and it has largely served in tandem with cognition. That is, rather than being the foe of emotion, reason has used emotion and emotions have exploited reason. Certain sorts of reasoning would be maladaptive without some emotional upsurge that halts thinking and forces action. The hominids who lingered to investigate whether the stripes glimpsed in the underbrush belonged to a predator didn't leave as many offspring as those who, driven by fear, simply fled at first glimpse. Emotions offer quick and dirty solutions to problems that make thinking risky. Alternatively, so-called commitment emotions may have evolved to
- strengthen group bonds, even if they work against self-centered rationality. Fathers have no rational reason to hang around after a woman is impregnated, but it seems
""likely that men who had romantic attachments to the mother had more children who survived, and so love helps unite father and mother across the lengthy period in which children grow to self-sufficiency.93 Within specific cultural contexts, of course, people learn to judge the proper moments to express feelings, to mask them with other feelings, or to send emotional signals.
In cinema, I'd suggest that emotion operates at all three of the levels I've sketched out. Most obviously, acts of appropriation are shot through with emotion. Fans cherish their favorite movies, critics get worked up in attacking a film they loathe, and unhappy viewers can wax indignant about a film's moral shortcomings. Less apparent are the ways in which emotions function in perception. A controversial case would be our startle response, which can be triggered quite automatically, as when you jump at a sudden burst of sound in a horror film. Startle isn't a prime candidate for being an emotion—it seems to prepare the way for the emotion of surprise—but it does lead to physiological arousal of a sort that primes affect.
More common and central is our sensitivity to emotional signals sent by other humans. Just as the rhesus macaques recognize signs of distress in their mates in a
52
Poetics of Cinema
movie, we are prepared to grasp many facial expressions- Newborn babies can reliably read their mothers' smiles, eye movements, and eyebrow play. A film's soundtrack can arouse us quite directly by cries, bellows, and other signals, just as infants respond to the mother's coos and baby talk.94 The weight of the evidence shows that evolution has primed us to engage in encounters with others by making us sensitive to the slightest signs of their emotional states. And these "affect programs" seem cross-cultural in large part, although we should expect them to vary extensively from person to person within any given culture.95 There is some evidence that in-group familiarity leads to faster recognition of facial expressions; Asians, whether living in Asia or the United States, recognize emotions on Chinese faces somewhat more readily than non-Asians do.96
More obvious are the emotions that fund comprehension. As we come to understand a narrative, we begin to run scenarios that require "emotional intelligence"— good guesses about how characters will react to the story situations. At the same time, we gauge a character's personality or current attitude on the basis of their emotional responses. Our inferential elaboration of the cues we're given is guided by the emotions that characters register. At the same time, the emotions we feel shape our sense of the film's macro-action. If we feel that a character has been wronged, we may mimic, in weakened form, her anger and self-righteousness. Screenwriters provide strong prompts for sympathy, such as making sure that the protagonist is treated unfairly, and many screenplay manuals argue that the skilful filmmaker evokes both hope (for the character's success) and fear (of the character's failure).97 Again, there may be considerable cross-cultural regularities in these emotions, most of which depend upon recurrent social situations that people in most cultures encounter—sympathy for children, anger at being wronged, and a sense of fairness or justice.
In comprehension, emotion and thought mesh, Greg Smith argues that narrative films tend to sustain moods and then punctuate them with bursts of emotion proper.98 These in turn can focus our attention on story developments. In Rope, our knowing where the corpse is hidden generates suspense, a mood that in turn makes us hyperattentive to every movement toward the chest. When Rupert lifts the lid, the mood prepares us to concentrate on his face, which betrays his shock and distress. Emotion also affects memory; in real life, a traumatic event becomes sharply etched into our minds. Films exploit this tendency by making the most vividly emotional scenes crucial for the plot—a death, a separation, a reunion. Ben Singer has proposed a catalogue of prototype scenarios, drawn from intense emotional experiences in ordinary human life, that melodramas draw upon. For example, the pathos we feel when seeing people degraded by misfortune forms the basis of a scene in Mother India (1957), when the mother, in the aftermath of a flood, digs desperately through the mud for something that will feed her dying children.99 Patrick Hogan's survey of transcultural story patterns traces their constant features to the way they make salient certain emotion-based prototypes of happiness.100
There's a great deal to be studied about how emotion works within our cinematic experience, including the bonding effects of watching a film with others. In some of the essays that follow, I propose that narrative films often model social intelligence,
Poetics of Cinema
53
Figure 1.22 The first shot of Ohayo (Ozu Yasujiro, 1959) shows a neighborhood and its clotheslines squatting under electrical towers.
Figure 1.23 Ohayo: The last shot presents the same configuration, but from the opposite angle. Ozu's typical reshuffling of shot elements is here applied to landscapes, emphasizing the washed underpants of a boy who can't control his bowels
and that modeling in turn rests upon sensitivity to emotional signals. Here all I want to indicate is that my last p, processing, can support systematic studies of the range of emotional effects that a film can have. More broadly, the cognitive perspective I've sketched here has informed several, quite different research programs, and many practitioners wouldn't endorse all the claims I've made.!0! My efforts are simply to show how a historical poetics of cinema of the sort exemplified in the essays that follow can usefully adopt some cognitive principles as a way to chart a film's manifold effects.
Sometimes, however, we can point to patterns and principles and purposes without being able to specify effects very well. This is largely because we don't have a tight theory allowing us to trace the consequences of artistic choices. Occasionally we may find constructional principles that don't correlate with any plausible effect. The begins/ring shot and final shot of Ozu's Ohayo (1959) are reverse-angle views of one another (Figures 1.22-1.23). Given Ozu's well-known interest in rejuggling objects and viewpoints and his use of 180-degree cutting patterns, it's likely that these two shots were planned to have the sort of regularity we find here.102 But when I point this out, someone will say quite reasonably that no one could perceive this pattern while viewing. Only by geeky concentration on Ohayo's overall architecture can we detect it. My response is that a few filmmakers build their films as objects as much as experiences, as patterned constructs that may or may not fit snugly in a viewing unfolding across time. Like a poet who plants hidden acrostics or numerical codes, Ozu expects some viewers to look at his film as if it were existing in a virtual space in which every shot can be compared with every other.
More generally, I'd suggest that we adopt the circuit particulars-patterns-purposes-principles-practkes-effects as a default to guide our inquiry. Most of the time it will serve us well. In a few cases, we may have to register the possibility that a film's organization can outrun its effects on any viewer. Yet by pointing out this state of affairs, the poetician may help the artist realize his or her design. Now that I've highlighted
54
Poetics of Cinema
the reverse-angle shots separated by an entire film, viewers are free to be alert for them when they see Ohayo, or other Ozu films. At this point, a connoisseur's appreciation for tiny felicities becomes the most relevant effect, and of course this too can be fruitfully investigated as part of an art tradition to which Ozu belongs.
Let me sum up. I've argued that broadly speaking, the central question of film poetics, posed as a methodological point of departure, can be understood in this way: How are films made in order to elicit certain effects?
'The first part of this formulation, dealing with the making, invites us to explore two domains. Analytical poetics studies the materials and forms of films to bring out the principles shaping them. Here we study theme and subject matter, large-scale form (such as narrative), and audiovisual style. Analytical poetics promotes functional explanations.
A second domain is historical poetics, the study of principles of filmmaking as they inform films in particular historical circumstances. This requires not only analysis of the films but also research into norms and craft practices impinging on the principles informing the films. It investigates how film artists, as historical agents, work within the zones of choice and control offered by their circumstances. Historical poetics thus traffics in both functional and causal explanations.
The last part of the initial formulation—the role of films in eliciting reactions-invites us to postulate that spectators play a role too. Call this a poetics of effect. Here we ask what activities are elicited by the thematic-formal-stylistic dynamics of the film and the principles undergirding same. I've suggested that studying viewing effects can fruitfully adopt a cognitive perspective, understood as involving perception, comprehension, and appropriation, all invested with emotion.
Overall, I've proposed a framework within which a variety of questions can be plausibly asked and answered. I hope it's clear that not every question that we pursue has to take all the particulars, patterns, and so forth into account. Sometimes just working on one or two components is quite fruitful. Nor is my anatomy an account of stages of inquiry. In studying any of these components, we are constantly moving back and forth among them. Sensitive critics have always shuttled between part and whole, material and form, form and function. I'm not urging that we create a rigid six-step procedure. I suggest only that we systematize our intuitions to a greater degree than usual, while allowing that historical analysis and recognition of the viewer's activity can enrich our sense of the film's constructive principles.
In stressing my own views, I've probably not done justice to the fact that poetics can host a variety of disparate research programs that can usefully debate alternative positions. For example, my solutions to certain problems will not be exactly compatible with those of the Russian formalists. Still, I owe to their great 1927 anthology, Poetika Kino, much more than the title of this collection. I share their broad theoretical ambitions and their methodological commitment to conducting rational and empirical inquiry into principles of art making within and across cultures. When I began writing from this perspective in the early 1980s, film studies was dominated by ideological critique and feminist psychoanalysis. From today's vantage point, I think that historical poetics, with its commitment to dialectical argumentation,
Poetics of Cinema
55
1
empirical research, and theoretical explicitness, has worn fairly well. Despite errors of fact, thought, and judgment—not least my own—the conversation has advanced. By concentrating on particular questions and then comparing our reasoning and research with that of others asking congruent questions, we have begun to produce reliable knowledge about film.103 A historical poetics of cinema isn't the only vehicle for this enterprise, but the essays that follow try to show that it can be a sturdy one.