The Classical Hollywood Cinema Film Style and Mode of Production to 1960
David Bordwell, Janet Staiger and Kristin Thompson
The classical Hollywood style, 1917-60
Neither normative criticisms nor morphological description alone will ever give us a theory of style. I do not know if such a theory is necessary; but if we want one we might do worse than approach artistic solutions in terms of those specifications which are taken for granted in a given period, and to list systematically, and even, if need be, pedantically, the priorities in the reconciliation of conflicting demands. Such a procedure will give us a new respect for the classical but will also open our minds to an appreciation of non-classical solutions representing entirely fresh discoveries. 1
An excessively obvious cinema
We all have a notion of the typical Hollywood film. The very label carries a set of expectations, often apparently obvious, about cinematic form and style. We can define that idea, test and ground those expectations, by using the concept of group style.
Historians routinely speak of group style in other arts: classicism or the Baroque in music, Impressionism or Cubism in painting, Symbolism or Imagism in poetry. 1 Cinema has its own group styles; German Expressionism, Soviet montage cinema, and the French New Wave afford timehonored instances. But to suggest that Hollywood cinema constitutes a group style seems more risky. In other national schools, a handful of filmmakers worked within sharply contained historical circumstances for only a few years. But Hollywood, as an extensive commercial enterprise, included hundreds of filmmakers and thousands of films, and it has existed for over six decades. If it is a daunting challenge to define a German Expressionist cinema or a Neorealist one, it might seem impossible to circumscribe a distinctive Hollywood 'group style.'
The historical arguments for the existence of such style are examined later in this book. At this point, a prima facie case for a 'classical Hollywood style' depends upon critically examining a body of films. Suppose that between 1917 and 1960 a distinct and homogeneous style has dominated American studio filmmaking-a style whose principles remain quite constant across decades, genres, studios, and personnel. My goal here is to identify, at several levels of generality, to what extent Hollywood filmmaking adheres to integral and limited stylistic conventions.
We could start with a description of the Hollywood style derived from Hollywood's own discourse, that enormous body of statements and assumptions to be found in trade journals, technical manuals, memoirs, and publicity handouts. We would find that the Hollywood cinema sees itself as bound by rules that set stringent limits on individual innovation; that telling a story is the basic formal concern, which makes the film studio resemble the monastery's scriptorium, the site of the transcription and transmission of countless narratives; that unity is a basic attribute of film form; that the Hollywood film purports to be 'realistic' in both an Aristotelian sense (truth to the probable) and a naturalistic one (truth to historical fact); that the Hollywood film strives to conceal its artifice through techniques of continuity and 'invisible' storytelling; that the film should be comprehensible and unambiguous; and that it possesses a fundamental emotional appeal that transcends class and nation. Reiterated tirelessly for at least seventy years, such precepts suggest that Hollywood practitioners recognized themselves as creating a distinct approach to film form and technique that we can justly label 'classical.'
We are not used to calling products of American mass culture 'classical' in any sense; the word apparently comes easier to the French speaker. As early as 1925, a French reviewer described Chaplin's Pay Day (1922) as a representative of 'cinematic classicism, ' and a year later Jean Renoir spoke of Chaplin, Lubitsch, and Clarence Brown as contributors to a 'classical cinema' of the future, one 'which owes nothing to tricks, where nothing is left to chance, where the smallest detail takes its place of importance in the overall psychological scheme of the film.' 2 It was probably André Bazin who gave the adjective the most currency; by 1939, Bazin declared, Hollywood filmmaking had acquired 'all the characteristics of a classical art.' 3 It seems proper to retain the term in English, since the principles which Hollywood claims as its own rely on notions
of decorum, proportion, formal harmony, respect for tradition, mimesis, self-effacing craftsmanship, and cool control of the perceiver's response-canons which critics in any medium usually call 'classical.'
To stress this collective and conserving aspect of Hollywood filmmaking also affords a useful counterweight to the individualist emphases of auteur criticism. Bazin criticized his protégés at Cahiers du cinéma by reminding them that the American cinema could not be reduced to an assembly of variegated creators, each armed with a personal vision: 4
What makes Hollywood so much better than anything else in the world is not only the quality of certain directors, but also the vitality and, in a certain sense, the excellence of a tradition…. The American cinema is a classical art, but why not then admire in it what is most admirable, i.e., not only the talent of this or that filmmaker, but the genius of the system, the richness of its ever-vigorous tradition, and its fertility when it comes into contact with new elements.
Bazin's point struck the Cahiers writers most forcefully only after his death, partly because the decline of the studio system faced them with mediocre works by such venerated filmmakers as Mann, Ray, and Cukor. 'We said, ' remarked Truffaut bitterly, 'that the American cinema pleases us, and its filmmakers are slaves; what if they were freed? And from the moment that they were freed, they made shitty films.' 5 Pierre Kast agreed: 'Better a good cinéma de salarie than a bad cinéma d'auteur.' 6 It is the cinéma de salarie, at least in its enduring aspects, that represents Hollywood's classicism.
All of which is not to say that Hollywood's classicism does not have disparate, even 'nonclassical' sources. Certainly the Hollywood style seeks effects that owe a good deal to, say, romantic music or nineteenth-century melodrama. Nor do Hollywood's own assumptions exhaustively account for its practice; the institution's discourse should not set our agenda for analysis. The point is simply that Hollywood films constitute a fairly coherent aesthetic tradition which sustains individual creation. For the purposes of this book, the label 'classicism' serves well because it swiftly conveys distinct aesthetic qualities (elegance, unity, rule-governed craftsmanship) and historical functions (Hollywood's role as the world's mainstream film style). Before there are auteurs, there are constraints; before there are deviations, there are norms.
Norms, paradigms, and standards
In the final analysis, we loved the American cinema because the films all resembled each other.
François Truffaut 7
The first, and crucial, step is to assume that classical filmmaking constitutes an aesthetic system that can characterize salient features of the individual work. The system cannot determine every minute detail of the work, but it isolates preferred practices and sets limits upon invention. The problem is, in other words, that of defining what Jan Mukařovský has called aesthetic norms.
When we think of a norm, especially in a legal sense, we tend to think of a codified and inflexible rule. While Mukařovský recognized that the aesthetic norms of a period are often felt by artists as constraints upon their freedom, he stressed the norms' comparative flexibility. He argued that the aesthetic norm is characterized by its non-practical nature; the only goal of the aesthetic norm is to permit art works to come into existence. This has important consequences: disobeying the aesthetic norm is not necessarily a negative act (may, indeed, be quite productive); and aesthetic norms can change rapidly and considerably. Mukařovský goes on to inventory several different kinds of norms, all of which intertwine within the art work. There are norms deriving from the materials of the art work. Poetry, for instance, takes language as its material, but language does not come raw to the task; it brings alongs norms of everyday usage. Secondly, there are technical norms, basic craft practices such as metrical schemes and genre conventions. Thirdly, there are practical, or sociopolitical norms; e.g., a character's ethical values represented in the work. Finally, Mukařovský speaks of aesthetic norms as such, which seem to be the basic principles of artistic construction that
form the work. These would include concepts of unity, decorum, novelty, and the like. 8
Mukařovský's work helps us move toward denying the Hollywood cinema as an aesthetic system. Plainly, the Hollywood style has functioned historically as a set of norms. It might seem rash to claim that Hollywood's norms have not drastically changed since around 1920, but Mukařovský points out that periods of 'classicism' tend toward harmony and stability. Moreover, the idea of multiple norms impinging upon the same work helps us see that it is unlikely that any Hollywood film will perfectly embody all norms: 'The interrelations among all these norms, which function as instruments for artistic devices, are too complex, too differentiated, and too unstable for the positive value of the work to be able to appear as virtually identical with the perfect fulfillment of all norms obtaining within it.' 9 No Hollywood film is the classical system; each is an 'unstable equilibrium' of classical norms.
Mukařovský's work also enables us to anticipate the particular norms which we will encounter. Evidently, classical cinema draws upon practical or ethico-socio-political norms; I shall mention these only when the particular ways of appropriating such norms are characteristic of the classical style. For example, heterosexual romance is one value in American society, but that value takes on an aesthetic function in the classical cinema (as, say, the typical motivation for the principal line of action). Material norms are also present in the cinema; when we speak of the 'theatrical' space of early films or of the Renaissance representation of the body as important for classical cinema, we are assuming that cinema has absorbed certain material norms from other media. Similarly, I will spend considerable time examining the technical norms of classical filmmaking, since to a large extent these pervasive and persistent conventions of form, technique, and genre constitute the Hollywood tradition. But in order to understand the underlying logic of the classical mode, we must also study how that mode deploys fundamental aesthetic norms. How, specifically, does Hollywood use such principles as unity and aesthetic function? As all these points indicate, the chief virtue of Mukařovský's work is to enable us to think of a group film style not as a monolith but as a complex system of specific forces in dynamic interaction.
My emphasis on norms should not be taken to imply an iron-clad technical formula imposed upon filmmakers. Any group style offers a range of alternatives. Classical filmmaking is not, strictly speaking, formulaic; there is always another way to do something. You can light a scene high- or low-key, you can pan or track, you can cut rapidly or seldom. A group style thus establishes what semiologists call a paradigm, a set of elements which can, according to rules, substitute for one another. Thinking of the classical style as a paradigm helps us retain a sense of the choices open to filmmakers within the tradition. At the same time, the style remains a unified system because the paradigm offers bounded alternatives. If you are a classical filmmaker, you cannot light a scene in such a way as to obscure the locale entirely (cf. Godard in Le gai savoir); you cannot pan or track without some narrative or generic motivation; you cannot make every shot one second long (cf. avant-garde works). Both the alternatives and the limitations of the style remain clear if we think of the paradigm as creating functional equivalents: a cut-in may replace a track-in, or color may replace lighting as a way to demarcate volumes, because each device fulfills the same role. Basic principles govern not only the elements in the paradigm but also the ways in which the elements may function.
Our account of this paradigm must also recognize how redundant it is. Not only are individual devices equivalent, but they often appear together. For instance, there are several cues for a flashback in a classical Hollywood film: pensive character attitude, close-up of face, slow dissolve, voice-over narration, sonic 'flashback, ' music. In any given case, several of these will be used together. In another mode of film practice, such as that of the European 'art cinema' of the 1960s, the same general paradigm governs a movement into flashback, but the conventional cues are not so redundant (e.g., pensive close-up but with no music or dissolve). The classical paradigm thus often lets the filmmaker choose how to be redundant, but seldom how redundant to be.
One more conception of Hollywood cinema as a unified system plays a part in understanding the classical style. This book will also refer to a
'standardized' film style. In general, this suggests only adherence to norms. But the term also implies that Hollywood cinema has been made stringently uniform by its dependence upon a specific economic mode of film production and consumption. Calling the Hollywood style 'standardized' often implies that norms have become recipes, routinely repeating a stereotyped product. Yet the avant-garde has no monopoly on quality, and violating a norm is not the only way to achieve aesthetic value. I assume that in any art, even those operating within a mass-production system, the art work can achieve value by modifying or skillfully obeying the premises of a dominant style.
Levels of generality
If the classical style is a set of norms, we need a way to distinguish greater and lesser degrees of abstraction in that set. A match-on-action cut is a classical convention; so is the principle of spatial continuity. But the first convention is a particular application of the second. Broadly speaking, we can analyze the classical Hollywood style at three levels.
1 Devices. Many isolated technical elements are characteristic of classical Hollywood cinema: three-point lighting, continuity editing, 'movie music, ' centered framings, dissolves, etc. Such devices are often what we think of as the 'Hollywood style' itself. Yet we cannot stop with simply inventorying these devices.
2 Systems. As members of a paradigm, technical devices achieve significance only when we understand their functions. A dissolve between scenes can convey the passage of time; but so can a cut. To say that the classical Hollywood style ceased to exist when most scenes were linked by cuts is to presume that a style is only the sum of its devices. A style consists not only of recurrent elements but of a set of functions and relations defined for them. These functions and relations are established by a system. For example, one cinematic system involves the construction of represented space. In classical filmmaking, lighting, sound, image composition, and editing all take as one task the articulation of space according to specific principles. It is this systematic quality that makes it possible for one device to do duty for another, or to repeat information conveyed by another. Thus employing a cut to link scenes conforms to one function defined by classical premises; within this paradigm, there must be some cue for a time lapse between scenes, and a cut may do duty for a dissolve (or a swish-pan, or a shot of a clock's moving hands). The systematic quality of film style also sets limits upon the paradigm; in representing space, for instance, ambiguous camera positions and discontinuous cutting are unlikely to occur because they violate certain principles of the system.
In this book, we shall assume that any fictional narrative film possesses three systems:
A system of narrative logic, which depends upon story events and causal relations and parallelisms among them;
A system of cinematic time; and
A system of cinematic space.
A given device may work within any or all of these systems, depending on the functions that the system assigns to the device. 10
3 Relations of systems. If systems are relations among elements, the total style can be defined as the relation of those systems to each other. Narrative logic, time, and space interact with one another. Does one of them subordinate the others? Do all three operate independently? How are the principles of one justified or challenged by another? In the Hollywood style, the systems do not play equal roles: space and time are almost invariably made vehicles for narrative causality. Moreover, specific principles govern that process. At this level, even irregularities in the various systems can be seen as purposeful. For instance, if we do find a passage of discontinuous cutting, we can ask whether it is still serving a narrative function (e.g., to convey a sudden, shocking event). In such a case, the relation among systems would remain consistent even if the individual device or system varied from normal usage.
We can, then, characterize the classical Hollywood style by its stylistic elements, by its stylistic systems, and, most abstractly, by the
relations it sets up among those systems. No single level of description will work. It is too narrow to define classical norms by devices, and it is unwarrantably broad to define them solely by relations among systems. (The domination of narrative logic over cinematic time and space is common to many styles.) Hence the importance of the second level, the stylistic systems. The categories of causality, time, and space enable us both to place individual devices within functional contexts and to see the classical style as a dynamic interplay of several principles. Finally, no categorical explanation of one level can wholly swallow up another. The systematic principle of depicting space unambiguously does not logically entail the use of three-point lighting. Those specific devices are the products of diverse historical processes; other elements might do as well. The specificity of the classical style depends upon all three levels of generality.
My account here will construct the classical stylistic paradigm across several decades, emphasizing the continuity at the second and third levels. But by stressing continuity of function I do not imply that the systems' paradigmatic range did not change somewhat. For example, before the mid-1920s, the use of high and low angles was severely codified: for long-shots (especially of landscapes), for optical point-of-view, or for shot/reverse-shot patterns when one person is higher than other. (In shot/ reverse-shot editing, an image of one element in the scene, typically a person talking, is followed by a shot of another element which is spatially opposite the first, typically, a person listening. Chapter 5 furnishes a more systematic explanation. See the examples in figs 16.65 and 16.66 in Chapter 16.) A medium-shot of an object or a human figure would seldom be framed from a sharp high- or low-angle. Yet in the late 1920s, Hollywood's spatial paradigm widened a bit, probably as a result of the influence of certain German films. Examples can be found in Bulldog Drummond (1929) and *The Show (1927), which dramatically use high and low angles (see figs 1.1 and 1.2). With the coming of sound, an occasional odd angle could compensate for what was felt to be an excessively 'theatrical' scene (see fig 1.3). Throughout the 1930s and 1940s, steep angles took their place as common functional equivalents for normal framings in many situations. Across history, the paradigm develops chiefly through changes in the first level of analysis-that of devices. This process will be examined in detail in Parts Three, Four, and Six.
Viewers, schemata, and mental sets
Considering the classical cinema as a system of norms operating at different levels of generality can seem to create a reified object, a colossal block of attributes that says little about how film viewers see films. The language of objectivism is hard to avoid, especially when we apply spatial metaphors like 'levels.' How, then, are we to characterize the viewer's work, or what E.H. Gombrich calls 'the beholder's share'?
An intricate and comprehensive theory of film viewing has yet to be constructed, and it is not within the scope of this book to do it. Yet if we want to consider how the Hollywood film solicits a specific way of being understood, we need to recognize at least how passive an 'illusionist' theory makes the spectator. Illusionist theorists usually insist that only avant-garde texts make the viewer perform an 'active' reading, or force the viewer to 'work to produce meaning.' 11 The Hollywood spectator, it is claimed, is little more than a receptacle; few skills of attention, memory, discrimination, inference-drawing, or hypothesistesting are required. Now this is clearly too simple. Classical films call forth activities on the part of the spectator. These activities may be highly standardized and comparatively easy to learn, but we cannot assume that they are simple.
Consider, as one problem, the spectator as perceiver. Illusionist theory emphasizes the deceptive quality of projected movement or of shot space: the spectator is duped into taking image for reality. As Noël Burch puts it, 'spectators experience the diegetic world as environment.' 12 But recent explorations in aesthetic perception and cognition have shown that 'illusion' is not simply a matter of fooling the eye. The spectator participates in creating the illusion. R.L. Gregory, for instance, speaks of perception as inferential, which makes 'illusion' dependent upon errors of inference: either biological 'mechanism' errors (e.g., the phi phenomenon as creating the illusion of movement) or cognitive 'strategy' errors (e.g., assuming that the whole is consistent with
displayed parts). Gombrich has also shown that visual illusion demands that the spectator propose, test, and discard perceptual hypotheses based on expectation and probability. 13 For illusion to work, the spectator must meet the art work at least half way.
If perceptual illusion requires some spectatorial activity, even more is required for that imaginative involvement solicited by narrative. No story tells all. Meir Sternberg characterizes following a tale as 'gap-filling, ' and just as we project motion on to a succession of frames, so we form hypotheses, make inferences, erect expectations, and draw conclusions about the film's characters and actions. 14 Again, the spectator must cooperate in fulfilling the film's form. It is clear that the protocols which control this activity derive from the system of norms operating in the classical style. For example, an insistence upon the primacy of narrative causality is a general feature of the classical system; the viewer translates this norm into a tacit strategy for spotting the work's unifying features, distinguishing significant information from 'noise, ' sorting the film's stimuli into the most comprehensive pattern.
Gombrich describes this process in terms of 'schemata' and 'mental sets.' Schemata are traditional formal patterns for rendering subject matter. Gombrich points out that the artist cannot simply copy reality; the artist can only render the model in terms of one schema or another. Thus even new shapes will be assimilated to categories which the artist has learned to handle. As Gombrich puts it, 'making precedes matching'-the creation of a schema precedes copying the model. 15 After the making, the schema can be modified in each particular case by the artist's purpose (usually, the sort of information the artist wants to convey). So far, much of this is congruent with Mukařovský's argument: we might think of the artist's schemata as technical norms and the artist's purpose as involving specific aesthetic norms. But Gombrich goes on to show that the schemata and the purpose function for the viewer as well. The artist's training is paralleled by the spectator's prior experience of the visual world and, especially, of other art works. The painter's traditional schemata constitute the basis of the viewer's expectations or mental set: 'A style, like a culture or climate of opinion, sets up a horizon of expectation, a mental set, which registers deviations and modifications with exaggerated sensitivity.' 16 For Gombrich, this mental set is defined in terms of probabilities: certain schemata are more likely to fit the data than others.
By pairing concepts like schemata and mental set, we can spell out the ways in which the classical film solicits the spectator. For instance, one well-known schema of Hollywood film editing is the shot/reverse-shot pattern. The filmmaker has this ready to hand for representing any two figures, groups, or objects within the same place. This schema can be fitted to many situations, whatever the differences of figure placement, camera height, lighting, or focus; whether the image is in widescreen ratio or not; whether the figures are facing one another or not; etc. Because of the tradition behind the schema, the viewer in turn expects to see the shot/reverse-shot figure, especially if the first shot of the combination appears. If the next shot does not obey the schema, the spectator then applies another, less probable, schema to the second shot. The spectator of the classical film thus riffles through the alternatives normalized by the style, from most to least likely. Through schemata, the style's norms not only impose their logic upon the material but also elicit particular activities from the viewer. The result is that in describing the classical system we are describing a set of operations that the viewer is expected to perform.
To stress the tasks which the film allots to the spectator allows us to abandon certain illusions of our own. We no longer need subscribe to copy theories of cinema, whereby a certain style simply replicates the real world or normal acts of perception; schemata, tied to historically defined purposes, always intervene to guide us in grasping the film. Nor need we imagine a Svengali cinema holding its audience in thrall. The classical schemata have created a mental set that still must be activated by and tested against any given film. Of course, the classical style defines certain spectatorial activities as salient, and the historical dominance of that style has so accustomed us to those activities that audiences may find other schemata more burdensome. Yet this dynamic concept of the viewer's role allows us to explain the very processes that seem so excessively obvious; as we shall see, even the
spectator's rapt absorption results from a hypothesis-checking that requires the viewer to meet the film halfway. We can also envision alternative viewing practices, other activities that the spectator might be asked to perform. The chapters that follow, then, suggest at several points how the norms of the classical Hollywood style encourage specific activities on the part of the spectator.
Style in history
If you're not working for Brezhnev Studio-Mosfilm, you are working for Nixon-Paramount. …You forget that this same master has been ordering the same film for fifty years.
Wind from the East
To construct the classical Hollywood style as a coherent system, we also need to account for the style's historical dimension. In one sense, this entire book tries to do that, by examining the Hollywood mode of production, the consolidation of the style in a specific period, and the changes that the style undergoes in subsequent years. At this point, I must indicate that my overall description of the classical style applies to a set of films across an extensive period. What historical assumptions underlie such a broadly based analysis?
The three levels of generality indicate some of those assumptions. My enterprise assumes a historical continuity at the two most abstract levels of style (systems and relations among systems); it assumes that the most distinct changes take place at the level of stylistic devices. For example, through its history Hollywood cinema seeks to represent events in a temporally continuous fashion; moreover, narrative logic has generally worked to motivate this temporal continuity. What changes through history are the various devices for representing temporal continuity such as inter-titles, cuts, irises, dissolves, whip-pans, and wipes.
By stressing the enduring principles of the classical style, we lose some specific detail. In this part, I shall not reconstruct the choices available to filmmakers at any given moment. If I say that a scene can begin by drawing back from a significant figure or object, that suggests that an iris, a cut, and a camera movement are all paradigmatic alternatives. But in 1917, the most probable choice would have been the iris; in 1925, the cut; in 1935, the camera movement. In discussing the general principles of classical style, I shall often project the historically variable devices on to the same plane to show their functional equivalence. This bird's-eye view enables us to map the basic and persistent features of the style in history. The more minute history of the devices themselves forms the bulk of Parts Three, Four, and Six.
Historical analysis demands a concept of periodization. Since we are concerned here with a stylistic history, we cannot presuppose that the periods used to write political or social history will demarcate the history of an art. That is, there is no immediate compulsion to define a 'cinema of the 1930s' as drastically different from that of 'the 1940s, ' or to distinguish pre-World War II Hollywood style from postwar Hollywood style. What, then, will constitute our grounds for periodization? Norms, yes; but also the film industry, the most proximate and pertinent institution for creating, regulating, and maintaining those norms. This is not to say that film style and mode of production march across decades in perfect synchronization. Parts Two and Five will provide a periodization for the Hollywood mode of production that while congruent in some respects, cannot be simply superimposed upon stylistic history. Nevertheless, we have chosen to frame our study within the years 1917-60.
The earlier date is easier to justify. Stylistically, from 1917 on, the classical model became dominant, in the sense that most American fiction films since that moment employed fundamentally similar narrative, temporal, and spatial systems. At the same time, the studio mode of production had become organized: detailed division of labor, the continuity script, and a hierarchical managerial system became the principal filmmaking procedures. Parts Two and Three detail how style and industry came to be so closely synchronized by 1917. But why halt an analysis of the classical Hollywood cinema in 1960?
The date triggers suspicion. Stylistically, there is no question that 'classical' films are still being made, as Part Seven will show. Variants of the Hollywood mode of production continue as well.
There are thus compelling reasons to claim that 1960 is a premature cutoff point. On the other hand, some critics may assert that this 'classical' period is far too roomy; one can see any period after 1929 as the 'breakdown' of the Hollywood cinema (the tensions of the Depression, the anguish of war and Cold War, and the competitive challenge of television).
The year 1960 was chosen for reasons of history and of convenience. In the film industry, it was widely believed that at the end of the decade Hollywood had reached the end of its mature existence. This Was Hollywood, the title of a 1960 book by publicist Beth Day, summarizes many reasons for considering the year as a turning point. Most production firms had converted their energies to television, the dominant massentertainment form since the mid-1950s; many had reduced their holdings in studio real estate; stars had become free agents; most producers had become independent; the B-film was virtually dead. 17 To Day's account we can add other signs of change. By 1960, a certain technological state of the art had been reached: high-definition color films, wide formats, and high-fidelity magnetic sound had set the standard of quality that continues today. Moreover, other styles began to challenge the dominance of classicism. The international art cinema, spearheaded by Ingmar Bergman, Akira Kurosawa, certain Italian directors, and the French New Wave, offered a more influential and widely disseminated alternative to Hollywood than had ever existed before. Not that Hollywood was significantly shaken (Part Seven tries to show why), but the force of the classical norm was reduced somewhat. Despite these reasons, it remains somewhat arbitrary to see 1960 as closing the classical period. We have chosen it partly because it makes our research somewhat manageable while still conveying the powerful spread of the classical cinema's authority.
The ordinary film
Film historians have not generally acknowledged the place of the typical work. In most film histories, masterworks and innovations rise monumentally out of a hazy terrain whose contours remain unknown. In other arts, however, the ordinary work is granted considerable importance. Academicism, mainstream works, the canon, tradition-the history of music, painting, and literature could not do without such conceptions. 'I believe, ' remarks Roman Jakobson, 'that a very important thing in analyzing trends in the cinema or the structure of a film, is the necessity of considering the base, the background of the spectator's habits. What films is the spectator used to seeing? To what forms is he accustomed?' 18 My analysis of the norms of the classical style thus gives privileged place not to the aberrant film that breaks or tests the rules but to the quietly conformist film that tries simply to follow them.
Between 1915 and 1960, at least fifteen thousand feature films were produced in America. It is impossible to analyze such a corpus. To construct a model of the ordinary film, we have selected in an unbiased fashion, 100 films from this period. (Appendix A explains the sampling procedures and lists the films.) We studied each film on a horizontal viewing machine, recording stylistic details of each shot and summarizing the film's action scene by scene. This body of data constitutes our unbiased sample (abbreviated UnS), and when we cite such a film, an asterisk signals it.
In the stylistic analysis of the cinema, the practice of unbiased sampling is unprecedented, but we believe it to be a sound way to determine historical norms. When the sample turned up what might be regarded as auteur films, we accepted this as inevitable in an unbiased sample and treated these films exactly as we did others. At least four-fifths of the sample, however, constitute a body of fairly obscure productions ranging across decades, studios, and genres. Furthermore, we have sought to test the conclusions about the UnS films by closely analyzing almost two hundred other Hollywood films of the 1915-60 period. We chose many of these films for their quality or historical influence, but many were as undistinguished as our UnS items. We shall refer to this second set of films as the Extended Sample (ES). My analysis of the classical style takes the UnS films as the central source of evidence and examples, drawing upon ES films occasionally. This means that many of the films mentioned will be unfamiliar to readers, but since I argue for their typicality, the
reader will recognize qualities present in many other films.
In one sense, the concept of group style simply makes manifest what we and Hollywood itself 'already know.' Concepts like norm, paradigm, stylistic alternatives, levels of systemic function, periodization, and schemata are, from this perspective, simply tools in making our habitual intuitions explicit. But these concepts also enable us to reveal the patterned and stable quality of our assumptions. The concepts can show that the classical cinema has an underlying logic which is not apparent from our common-sense reflection upon the films or from Hollywood's own discourse about them. The theoretical concepts introduced in this chapter are indispensible to grasping the classical style's systematic quality. Armed with them, we can go on to examine how that style characteristically organizes causality, time, and space. The next five chapters, then, should trigger a certain déja vu; the reader will recognize some familiar filmmaking practices. But these chapters also seek to explain in a systematic way how these practices work together to create a distinct film style which, like Poe's purloined letter, 'escapes observation by dint of being excessively obvious.' 19