Gunning - Presence of Narrator: Griffith's Films PDF

# Presence of the Narrator: The Heritage of the Films ## Biograph of Griffith >> D.W. Griffith **Thomas GUNNING** _Colloquium International, Paris, Eclitibus/Harmattan_, _1984, p. 126-147_. _Car un père est Arthur Rimband_ The re-writing of the history of cinema that each generation undertakes...

# Presence of the Narrator: The Heritage of the Films ## Biograph of Griffith >> D.W. Griffith **Thomas GUNNING** _Colloquium International, Paris, Eclitibus/Harmattan_, _1984, p. 126-147_. _Car un père est Arthur Rimband_ The re-writing of the history of cinema that each generation undertakes encounters the same obstacles presenting them- selves in different forms. If the model of cinema history that con- sisted of chronicling technical and aesthetic progress (from short films to longer films, from silent films to sound films, from black and white films to color films) is outdated, it is perhaps more difficult to distinguish oneself from a history of film style that ultimately prioritizes the styles that have dominated. This is apparent in the work of André Gau- dreault and Charles Muser on the films of Edwin Porter (especially "The Life of an American Fireman") (1). In their haste to find the origins of continuous editing, previous historians neglected the versions of this film that included editing with temporal overlap (in which the action is repe- ated after the "match cut") because this style of editing is aberrant from the point of view of later techniques. It is indispensable to undertake a history of cinema that takes into account discontinuities in the development of style, fissures in the apparent monolithism of film practice, detours and blind alleys in the course of the evolution of film. This is particularly true for the study of the early films of D.W. Griffith. The ambiguous designation of Griffith as the "necessary patriarch" of classical narrative cinema (2) must be accompanied by a precise study of this claim of paternity. Many aspects of this research fall outside the scope of my communication: the multiple sources of narrativity in film; the many creators of films who contributed to the elaboration of the classical style; the effect and the scope of Griffith's direct influence. But attentive viewing and precise analysis of the corpus of Griffith's early films make it possible to address a fun- damental question: the definition of the narrative style in these early films, an understanding of the nature of Griffith's work that became the heritage of later film creators. I want to show that this heritage of Griffith is more problematic than is often supposed. In a penetrating essay, Jacques Aumont has already challenged the notion of positing Griffith as the source and origin of the classical narrative style, the style "of transparency and naturalization" (3). A broader defini- tion of this classical narrative style (especially in rela- tion to editing) is given by Alain Bergala: "Classical narrative editing (in cinema, for example) is deter- mined by an ideology of fiction, of which we have spoken before, which is that of the impression of reality, transparency. To meet the requirements of this ideology, the world of fiction must present itself as a "natural" world, continuous and homogeneous, and must make itself forgotten for what it really is, that is, an imaginary production whose materiel is always fragmented and discontinuous" (4). My goal here is to place the analysis of J. Aumont in a historical context and to focus my intervention directly on that aspect of Griffith's style that is most often considered his heritage: the development of editing and more particularly of intercut editing. In redefining the heritage of Griffith, it is not my inten- tion to deny the impact of his style on the develop- ment of the classical narrative style of cinema. As I have indicated in another work (5), the style of the early films of Griffith develops in a context of economic and ideological transformation of the Ame- rican film industry from which classical narrative film will be defined for the first time. This is clear- ly apparent in the formation, in 1908, of the Motion Picture Patents Company, an organization whose avowed economic policy was to try to raise the social respectability of cinema and therefore to attract the midddle class. The evolution of film style that takes place during Griffith's presence at Biograph must be considered in part a creation of certain filmic codes that were capable of effectively translating the values of the bourgeois ideology already being véhicu- lated by theater and contemporary novels. The important role played by Griffith in this transforma- tion must always be situated in the context of an evolution of the social role of cinema in the United States during this period, an evolution that is clear- ly analyzed in the specialized journals of the time. The aesthetics and the ideology of this change are brilliantly summarized in the "cinema" column begun in 1909 by the professional journal "The New York Dramatic Mirror." This weekly column was written by a man named Frank Woods under the pseudonym "The Spectator" (6). We find here, clearly stated, the ideal that will later be identified as that of the classic diegesis: a transparent fiction whose impression of reality is strong enough to conceal from the spectator any trace of the process involved in its fabrication. I will sum- marize and quote Woods's commentary to illustrate how this ideal, according to him, should be véhicu- lated by films of the early era. The commentary of Woods indicates with what awareness the foundations of the classical style were being laid. However, as I will indicate later, I do not believe that Woods' comments should be interpreted too simply. First, the very fact of stating the need for transparent illusionism (compared to later periods in the history of cinema where this need was taken for granted and almost forgotten by film creators, critics and the public) proves the precarious status of this ideal during the birth of narrative film. Second, as I will show by studying the films of Griffith directly and in particular, their style of editing, the ideal of transparent narration is only one element in the narrative style of Griffith. The first expression, in Woods's column, of the need for an effacement of the work of production appears in his critique of the habit that film actors had of appearing conscious of the presence of the camera (a habit that was commonplace in the films of Porter, Méliès, Zecca, etc.) Woods summarizes his objections concerning this practice: > "There is an important failing in almost every ac- tor, one that's underestimated, a tendency that's common to almost every actor to appear conscious of the presence of the camera... A good director constantly insists on his actors to turn their eyes away from the camera, and good actors try to do that. Many succeed; but simply not looking at the camera is not enough? Shouldn't there be an absolute indifference to the presence of the camera? Of course, we admit that the director and the actors should never forget the camera. Every action must occur in a way that the camera can get the best possible view of the scene, but isn't it true that the closer actors can seem to be unaware of the presence of the camera, the closer they will come /to achieving absolute realism? As it has already been explained in this article, acting in film is more effective if it creates the illusion that it represents real events and not made-up events (7)." This critique of the awareness that an actor can have of the presence of the camera was based on Woods's conviction that the very special power of cinema resulted from its amazing ability to pro- duce an illusion of reality. An illusion that gave cinema a "peculiar" psychological power over its public: > "The peculiar power that attracts you to films resides in the impression of reality that films convey; by means of this impression of reality, films exert an influence on the mind of the spectator similar to hypnosis or suggestion by visual means; this kind of limited hypnotic influence is capable of more powerful effects through cinema than through any theatrical or literary production, and it is therefore wise to cultivate absolute realism in films in all areas of cinematic art. Artificial drama or comedy seem to have no effect on the mind of the audience when they are adapted for cinema, whatever qual- ity they may have displayed on stage or in writ- ten literature (8)." Any deviation from this aesthetics of vraisemblance, for example, an actor betraying the presence of the camera, could destroy this hypnotic effect: > "By turning his face towards the camera, he betrays the fact that he is acting, shows that there is some- one in front of him, hidden from the spectator, to whom the actor is addressing himself. Immediately, the impression of reality disappears and the hypnotic illusion that had taken hold of the spectator through the power of visual suggestion vanishes. It's like if a hypnotist snapped his fingers in front of his subjec- t and said 'Wake up!' The cinema audience neither realizes nor analyzes this shock, but they feel it, and the subtle charm of the action is effaced (9)." The illusion of reality that Woods describes is that of a closed-circuit diegesis that does not appear to refer to any source of production. Any awareness of the camera on the part of the actor, or awareness of being a spectator on the part of the viewer, would weaken this hypnotic illusion, the first implying the second. Woods's aesthetics is based on a denial of both the camera on the one hand, and, on the other, the recognition of the spectator's act of viewing the film: "we, spectators, are not part of the image, any more than a camera is supposed to be present" (10). In the years that followed, Woods concern for the maintenance of a unified fictional world would extend to the principles of editing. In 1911 we find him inter- rogating the principle of directional continuity in classi- cal editing: > "Now we are examining the question of the fusion of scenes - the need for judicious and skillful judgment in jumping from one scene to another... (one should avoid) the bad habit that some directors still have of ignoring directions when moving from one scene to the next... We see a person enter a house following a given direction; in the next scene, while he is still completing his movement of entry inside, he appears to have simply jumped to the other side of the room and is now moving in the opposite direction. It's as if the reader of a book were suddenly faced with characters written in reverse and had to turn the book around to continue reading. This is in effect what one must do in thought when scenes like the one described above are projected on the screen - one must mentally jump to the other side of the house. The spectator refuses to make this jump (11)." These are the foundations of the system of conti- nuous editing in which the spatial orientation of each shot is coherent with that of the other shots in a sequence. Space and the action appear as a unified whole despite their fragmented presentation in each shot. Woods is asking here for a logical orien- tation of the spectator relative to the projected image so that right and left maintain a constant relationship, regardless of the cut. This constancy of orientation (that Woods interestingly compares to the relationship of the reader to the written text) will lay the foundation for what is to become the 180- degree rule of continuous editing. This search for a sense of the continuity of the action and of space above its fragmentation through editing constitutes the principle of clas- sical editing, the style described by Marie-Claire Ropars-Wuilleumier as that which "always operates on matter conceived as continuous representation, has a purely analytical function: meant to make things seen better, by eliminating the unnecessary and by emphasizing the essential, it must also make itself forgotten, by maintaining the illusion of continuity in space, time and action" (12). Woods praised a Biograph film of Griffith's (unfor- tunately one of the 60 or so Biograph films directed by Griffith that are not currently available, although all indications are that this film would be typical of Griffith's editing in 1911), "The White Rose of the Wild" for its ability to "assemble scenes": > "To an experienced eye, one of the most striking features of this film is the almost perfect mechanic- al precision with which each scene is timed: every- thing unfolds like well-oiled clockwork, without inconsistency. In several instances, when a scene joins another at a point where an entrance is made through a door, and instantaneously we see the movement complete on the other side of the door, the action is so carefully articulated that it seems to be one single movement. This approaches perfection in the technique of staging (13)." This text of Frank Woods gives us a very clear formulation of the ideology of the transforma- tion of film style that occurs in the United States between roughly 1908 and 1913, and in which Griffith plays such an important role. Woods's column announces many of the principles of clas- sical style: its illusion of realism; its transparency; its effacement of the traces of production; its effort to guide the spectator automatically and harmoni- ously; its creation of a continuous and homogeneous diegesis. The role of Griffith in establishing the ideology of the classical narrative film style having been established, why do I claim that this does not account totally for the heritage of Griffith? What in in the staging style of Griffith's films contradicts (or is obscured by) this understanding of classical style? I am convinced that the act of narration in Griff- ith's very early films - the filmic codes that are introduced or become habitual during Griffith's career at Biograph and that make possible the development of film as a narrative form - is not based simply on the classical conception of continu- ity. The aesthetics represented by Woods is the means by which the fundamental nature of Grif- fith's narrative technique is rehabilitated. This rehabilitation (and the narrative form that it makes possible) is an essential part of Griffith's heritage, but only one part. This should not lead us to lose sight of the fact that the act of narration in Grif- fith's early films is based on discontinuity: the discovery of the "cut" as a means of transforming the shot and of elevating it to a new degree of significance. What I am saying here is obvious in a certain sense. Before continuous editing could appear, we had to have the fragmentation of the "cut." Before sutureing the diegesis, we have to make an inci- sion. Of course, our conception of the classical style of editing implies this. The style of classical Holly- wood narrative film is founded on the mastery or concealment of the discontinuity inherent in editing (as Bergala writes, "the first contradiction that the code of editing must resolve is therefore that of con- structing the illusion of a continuous and homo- geneous universe from fundamentally fragmen- ted and discontinuous materiel") (14). However, I propose a shift in perspective, Recent theorists of film have described how this discontinuity is reabsorbed by the continuity of a homogeneous diegesis. To really understand the early films of Griffith (and the beginnings of narrative codes in cinema) one must look at the period that predates this re-establishment of continuity: the tendency toward fragmentation, the decision to cut. This shift in perspective implies a historical point of view that takes into account the difference between the period when Griffith created his films and the decades that followed in which the rules of editing were established, governing the system of continuous editing. After the establishment of the rules of continuous editing in Hollywood during the 1920s and at the beginning of the 1930s, the rules that would determine the logic of the editing process would also preordain the logic of camera positions during shooting, offering filmmakers a limited choice of possible positions. In Griffith's Biograph films (and even in his later films) the rules of contin- uous editing did not yet determine the process of shooting and they did not yet have a place in editing. The fundamental discontinuity between shots is more evident and the processes of shooting and of editing are more in opposition than in later years when the rules of continuous editing would determine and coordinate both processes. While Griffith's style of editing unites shots into a unified diegesis, the joint is visible, attention is drawn to it. The cut is exhibited at the same time as it is sut- ured. As Jacques Aumont puts it, it is "sewn with white thread" (15). But to place the style of Griffith's early films historically, they must be seen in the context of the films that preceded them, rather than those that fol- lowed them. I will try to look back quickly and schematically (and incompletely in a few lines) at the history of the idea of continuity in films before Griffith. In 1900 and 1907 approximately, we find a large number of film practices that appear anomalous relative to the concepts of continuity in later films (16). I have already mentioned two of the most important, the awareness of the presence of the camera in actors and the repetition of action with temporal overlap, a "match cut" where time seems to tread water rather than progress continu- ously. It's also important to note that before 1904 (at least in the United States), cinematic narratives generally included only one shot, films with more than one shot being a complete novelty at the time. In fact, multi-shot narratives were generally, as has often been noted, just a series of tableaux illustrat- ing a certain narrative progression. Each tableau had a relative autonomy and generally contained an action unfolding entirely within a given frame. There was little syntactic interdependence between the different tableaux outside of a certain narrative logic. It was often necessary to know the materia of inspiration of the film in advance or to listen very carefully to the spoken commentary accompanying the screening to make the connection between these practically independent tableaux and make them into a coherent narrative (see, in this regard, the series of Biograph films inspired by a popular melodrama about alcoholism entitled "Ten Nights in a Barroom" where the spectator is shown only the key events in the narrative; or Porter's "Uncle Tom's Cabin" whose narrative arrangement is so disorganized that it took until the present day, to see the version distributed by the Museum of Modern Art in New York, to realize the disorder in the narrative). The immense popularity of "chase films" from the year 1904 onwards must be partly explained by the fact that these films succeeded in creating a certain continuity of the different shots of which they were composed. These stories of pursuit allowed the establishment in a very simple manner of a syntactic relationship between separate shots and to create a kind of spatial continuum that transcended the frame of each shot. From now on, a series of successive shots had to represent the trajectory of a single action - the pursuit - and, while each shot took place in a new setting, this trajectory linked the different settings in a continuous geog- raphy. The scheme of these chase films (from "The Escaped Lunatic," filmed at Biograph in 1904, to Griffith's "Balked at the Altar" - 1908) remains pretty similar throughout the period when this kind of story was in vogue. Pursuers and the pursued are always shown in the same shot; there is no editing process that separates them. Thus, whether it's a lunatic on the run, a potential fiancé or a chicken thief, they are always shown entering the frame and then followed a little further by the asylum workers, the hordes of old maids or the angry farmers, as the case may be. And, most often, the shot ends when all the moving characters have left the frame (this is particularly the case of the very early films of this genre). The characters thus leave the frame to disappear momentarily and then re- appear, as if by magic, in the following shot exactly in the order in which they were left, but in a totally different setting. This genre is also a kind of cinematic equivalent of the "Fort/Da" game described by Freud in "Beyond the Pleasure Princi- ple" in which the observer takes pleasure in seeing objects disappear and then reappear (17). We will call "trajectory continuity" this primitive form of continuity that follows the trajectory of a character or a group of characters through dif- ferent sets that make up a "continuous geography." This formula was quickly adapted to a whole range of narratives, particularly to variations on chase films such as Hepworth's "Rescued by Rover"(1905), and to a certain type of comedy that we have called "linked sketches." In these comedies (particularly popular between 1906 and 1908), a character faces a number of situations along a unique path ("A Lady Truly Fair" by Louis Feuillade, Gaumont, 1908, in which a pretty woman walks down a street in Paris and causes a series of dis- asters along the way in attracting the attention of the men she encounters, is a good example of this then international genre). Griffith inherited this form of continuity and used it in a number of his early films made for Biograph, such as "Balked at the Altar" and "The Curtain Pole" (both made in 1908). In fact Griffith's very first film "The Adventures of Dollie," 1908, already uses this type of continuity. In this system of "trajectory continuity," the "match cut" between two shots serves to link two phases of a character's movement and artificially creates continuity and progression in the trajectory by concealing the fundamentally elliptical nature of the genre. The "match cuts" thus link two distinct phases of the character's trajectory by satisfying the promise of a shot ("The pursuit continues...") by the arrival of the following shot, and by contin- uing this process to the end, that is until the com- plete series of successive shots is finished because of the end of the character's trajectory. Griffith used this form of continuity to make it the basic structure of several films or a simple element in the general construction of many other films. But he also, within this formula, created a tech- nique of discontinuity that would ultimately change the very nature of cinematic narration. In this new style of editing, the "match cut" does not simply serve to satisfy the viewer's expectation by showing him the next stage of the action, but goes so far as to use its power of interruption of an action. From now on, the "cut" intervenes in the natural unfolding of an action, interrupts its progression and ultimately hangs it up. The type of editing that we are referring to will later be known as intercut editing. The history of this procedure, while seeming relatively familiar, has many blind spots. We know in any case that this technique existed before Griffith. But we also know that some of the first examples that are usually given as constituting the beginnings of intercut editing, such as Williamson's "Attack on a Chinese Mission" (1900) or "The Great Train Robbery," by Porter (1903) -- are only ambiguous prototypes (18). In fact, for films before 1908, there are few cases of prolonged parallelism between two actions taking place in different settings (19). Generally, this type of narrative device is very rare before 1908 and only really gains broad development at the time of Griffith's first films. The increasing use of this purely cinematic narrative device (20), observable with Griffith, will entail a number of important trans- formations in cinematic language, particularly the determination of spatio-temporal relation- ships between the different shots of a film to a degree that was not necessary in the procedure called "trajectory continuity" (21). But what we are interested in analyzing more particularly here is the interweaving of two different actions on the basis of a constant interruption of each of these actions and a movement from one to the other. As we have already pointed out, the type of editing that is observed in films with "trajectory contin- uity" links two phases of the same action. In contrast, in "intercut" editing. the progression of a certain action is interrupted by a shot of a different action, totally different even if it is part of the general framework of the narrative. In his early uses of intercut editing, Griffith simply moves from one action to the other (in this case, the second action often served as a counterpoint to the first). For example, in "The Greaser's Gauntlet" (1908, the first case of the use of intercut editing by Griffith), the narrative moves from a shot of a lynch mob that has seized the innocent José to a shot of his lawyer, Mildred, who is trying to find the real culprit. Even in this very simple form, this type of editing creates a new relationship with the spectator, wholly different from that established by "trajectory continuity." In intercut editing, the blending of two different actions requires the spectator to hold his expectation for the first action in order to follow a second action, and this process delays the almost mag- ical reappearance of the first movement that has been temporarily interrupted. This effect produced on the audience has been aptly named "suspense." But the more that Griffith uses intercut editing and masters it, the more he structures it to accentuate as much as possible the factors of interruption in a shot. This is in total opposition to the trend in the early "chase films" to keep a shot going until the last figure in movement is entirely out of the frame and to cut only when all the possi- bilities for narrative progression have been exhaust- ed. But by the end of 1908, Griffith was already beginning to place his "cut" at the moment of maximum narrative intensity of the shot and so as really to interrupt the natural unfolding of the action. Thus, the "cut" gained increasing impor- tance and size in its role of interrupting and reorganizing the action. One of the first examples of this"cut" is found in the justly celebrated cinematic adaptation of Enoch Arden made by Griffith under the title "After Many Years": > "The editing of this film interweaves two actions: that of the sailor John Davis who has been ship- wrecked, and that of his wife who waits for news of him in their home. The camera goes several times from the desert island where John has taken refuge to the patient wife somewhere in England (shots 3, 4, 5 and 6), then comes an extraordinary "match cut" linking shots 8 and 9. In shot 8, we see John standing in front of his tropical hut; he has in his hands a small box that we assume contains a portrait of his wife; John opens the box and takes out a portrait of his wife that he kisses. Then, right in the middle of this action, Griffith cuts and goes to a shot of the wife waiting on the porch of their English cottage. In fact (just like the case of the false reverse-angle shot observed by Aumont in a quasi-remake of this film, entitled 'Enoch Arden' and dating from 1911), the action is taken up to a theoretically impossible denouement since it ignores very large geographical distances: the wife actually receives the kiss of her distant husband and opens her arms as if he were right in front of her. We can say that if there is a kind of metaphorical (and perhaps even metaphysical) reunion of the couple here, it is founded on the interruption of the action unfolding in the first shot. The fact of leaving this kiss in suspense in shot 8 (in total opposition to the editing style of the early "chase films") undermines the impres- sion of completeness and autonomy that this shot could give. Thus, this shot is interrupted at a certain point in order to function as a fragment that only finds its complete meaning when placed within a set of shots. This way of interrupting action (which is opposed to the technique that will later be used in Hollywood, a technique tending toward mak- ing editing imperceptible by smoothly moving, for example, the movement of an actor from a long shot to a medium shot, to mask the "break" that the "cut" represents) will become more and more important with Griffith through his incre- asingly frequent use of intercut editing, in sequences such as, for example, in the melodramatic build-up towards the resolution of a desperate situation. In these sequences, as we have seen, one of the roles of intercut editing is to "delay" the narration and to suspend the solution to a dramatic situation: thus in "The Greaser's Gauntlet" that we were discussing above, we wonder if the lynch mob will hang José before Mildred arrives with the proof of his innocence. Or, in another case, will the trap by which a pistol is connected to a clock be triggered before the rescuers that might free the detective held prisoner in the face of this deadly weapon arrive? But in fact Griffith accentuates the suspense by interrupting the key gesture. The first example of this type of interrup- tion of action is given to us in "The Barbarian Ingomar" (1908). In shot 14 of this film, we see a young Greek woman, Parthena, who has been tied to a tree by a band of barbarians. Then one of the renegades raises his sword to kill the girl. At precisely this moment, Griffith inter- rupts the murderous gesture and goes to a shot of Ingomar rushing towards the barbarians' camp to save Parthena. Then shot 16 brings us back to the camp and the fatal sword, which still hangs over the girl, until Ingomar enters the frame and comes to slay the would-be assassin. The longest sequence using this type of suspense- building montage appears in "The Drive for Life" (1909). This is a melodrama in which an aban- doned mistress has sent a box of poisoned cho- colates to her ex-lover's fiancée. The climax sequence of the film first shows us the fiancé who has just learned of his former mistress's act of vengeance, then goes to the fiancée receiving the poisoned sweets and preparing to eat them. The central shots of this sequence cut the action up into suspenseful montage: > **Shot 23:** A general shot of a parlor. Mignon, the new fiancé, sits reading, next to her mother and sisters. A valet enters the room accompanied by a small delivery boy who brings the box of chocolates. Mignon takes the box and reads the card that accompanies it (and that we can assume indicates the giver of the gift), then kisses the card and starts to open the package. The action is interrupted at this point for **Shot 24:** A suburban road, with a series of houses on the right. A car with Harry Walker - Mignon's fiancé, who has gotten wind of the poisoned chocolates - is driving towards the camera, then leaves the frame. **Shot 25:** A country road. Harry Walker's car arrives towards the camera, then stops. Walker and his driver get out and examine the engine. The two men get back in the car after making a minor repair, then the car restars, still headed towards the camera, until it leaves the frame. **Shot 26:** The parlor in Mignon's house (as in shot 23). The girl opens the box of cho- colates and takes one out. She brings it to her lips. Cut to: **Shot 27:** A country road, photographed from a slightly oblique angle. Appearance of Walker's car emerging from the right. **Shot 28:** A private gate on a country road. It is open to let a horse-drawn carriage pass, then closes behind the carriage. A man comes out of the frame. It is at this point that Walker's car appears in the background and approaches the camera. It passes the gate and exits the frame. The gatekeeper returns to the frame shouting after this intruding car (which has already left the screen) and shaking his fist in anger. **Shot 29:** Back to the parlor (the same shot as 23 and 26). Mignon takes a chocolate, brings it to her mouth, but only kisses it (as she did earlier with the card) before put- ting it back in the box. Mignon sits back down next to her sisters, one of whom is about to take a chocolate. But in fact, she hands it to Mignon who opens her mouth to swallow it. At this point, cut to: **Shot 30:** A country lane. The horse-drawn carriage stops at the side of the road; the wheels are on the shoulder. Walker's automo- bile comes at high speed, hits the carriage and keeps going. **Shot 31:** A slightly different angle of the same setting. Men standing next to the damaged horse-drawn carriage are shouting after the Walker car (which is off-screen). **Shot 32:** Another stretch of road. Walker's car approaches the camera, rounding a ben- d **Shot 33:** The parlor. Mignon's sister still holds the chocolate suspended over Mignon's open mouth. Another sister tickles her, and the first sister drops the chocolate. They both bend down to pick it up. The interruption of these different actions in this sequence functions as a series of delays in the resolution of the narrative, following a tradition of melodrama that is already ancient. However, we can say that there are essentially two types of narrative "delays" in this sequence. On the one hand, there is a series of delays within the diegesis itself - the repair of the engine; the first gate; the horse-drawn carriage; the kiss on the chocolate at the moment when we think Mignon is going to eat it; the fall of the chocolate to the floor. But, on the other hand, there are also the delays provoked solely by the style of editing itself, which structures and suspends the action of the different characters. The stopping of the shot where one of the sis- ters is about to place the chocolate in Mignon's open mouth (an action interrupted between shots 29 and 33) is just the most extreme example of this manipulation of action by editing, which structures it and delays its reso- lution (22). We are here in the presence of a style of narra- tion that organizes the different actions contained within each shot by editing. Some might argue that in a film, any recourse to editing, any system of organization that goes beyond the autonomy of each shot, entails this type of narration. But in reality, the editing style that characterizes the early films with "trajectory continuity" is entirely tied to the action contained within each shot. The shots (and the settings) form a kind of chain resting on the movement of the characters in each shot. In contrast, in "The Drive For a Life", the style of ed- iting and of narration is based on a principle of independence relative to the action inscribed in each shot, in such a way that the "cut" can very well interrupt an action and suspend its resolution. This new system of narration also creates a new relationship with the spectator. The editing of the early chase films led to a narrative style that immediately responded to the audience's expectation (a character exits the frame, but quickly reappears in the following shot). In contrast, the suspense effect created by the type of editing that is observed in "The Drive For a Life" derives precisely from the fact that it delays the satisfac- tion of this spectator's expectation; for it tends to suspend the resolution of the action, often, besides, for the duration of several shots. Perhaps one might relate this new mode of nar