Repeated long takes of the same narrative moment
Corridor encounter: first version. Elias photographs John in the corridor while Michelle runs past.
Corridor encounter 2 shows the same action, seen in 180 degree reverse angle.
Corridor encounter 3: the moving camera follows Michelle as she moves past Elias and John.
Moments when the narrative witholds information we want to know
At the end of the film, Alex corners Nathan and Carrie in the freezer. We do not know if he shoots them or what happens to him.
Michelle hears the sound of a gun being cocked and looks off screen...
... but instead of a corresponding POV shot, Van Sant cuts to a completely different scene, leaving Michelle's glance "unanswered."
A visual generation gap: John and his father
Father and son are in the same space but....
...separated by framing.
Photographing two people sitting in a car is common, but Van Sant makes a special effort to keep the generations separated.
Visual generation gap between students and administrators
Nathan and Carrie talk to an off-screen adminstrator.
John talks to an unidentified administrator.
As Michelle walks from the athletic field to the gym, a PE teacher scolds her, screen right, and drifts in and out of a tracking shot that favors Michelle.
Visual generation gap:
Eric has breakfast with faceless parents.
Alex and unseen delivery man, delivering the gun.
Teacher and student unified by framing, separated by focus.
Michelle, isolated visually and emotionally, endures the ridicule of partially viewed classmates.
A continuous shot reveals separate worlds. The teacher answers a question, the camera pans left and...
... the shot continues. Jock bullies unseen student. Camera continues pan left to reveal...
... Alex covered in spitballs, tormented by now off-screen students.
The craftsmanship of photography
Elias cuts film the old-fashioned way, as ...
... Van Sant edits film the old-fashioned way, on a flatbed instead of computer, as shown here in a behind-the-scenes documentary.
Although in Gerry and Elephant Van Sant employs Bazinian principles more devotedly than most contemporary American filmmakers he also undermines the claim to realism these principle evoke. He revives the principles and demonstrates their limitation at the same time. Bazin advises,
Van Sant is not so circumspect. On the school athletic field, before Nathan begins his long walk, a football game is in progress. Instead of panning to follow the action the camera remains stationary and the game migrates in and out of the frame. The choice of what to see is partially ceded from the camera operator to the actors.
A more complex dialectic of inclusion-exclusion is created in a meeting of the Gay-Straight Alliance (GSA). A teacher and several students sit in a circle with the camera in the center. The camera pans/rotates steadily from one participant to another in a continuous left to right movement while people talk. But the person speaking isn't always the person onscreen. We hear people talking without seeing their faces and see people just listening. When someone starts talking the camera doesn't pan or cut to them in the manner of a typical documentary interview, but just keeps panning. We have to wait to be able to associate each voice with a face, which we can only partially accomplish. Whether we actually see the person talking is a matter of chance. No matter how much activity a wide angle lens and a mobile camera take in, activity is still excluded. What the camera sees, rather than being “obvious” or “natural,” can be arbitrary (Ray 286).
If the camera offers a rich but a limited point of view, how can a filmmaker depict a complex event such as the Columbine shootings? The traditional answer—innovated in simple movies like The Great Train Robbery (Edwin S. Porter, 1903)—is parallel montage: cutting back-and-forth between two or more sets of activity that the viewer intuits to be occurring simultaneously. D.W. Griffith perfected this technique in the suspenseful climaxes of films such as The Birth of a Nation (1915), Way Down East (1920), and America (1924) in which trapped, terrified women are intercut with men racing to their rescue. Eisenstein used this technique for political as well as melodramatic ends in films such as Strike (1924) and The Battleship Potemkin (1925) in which violent conflicts involving large crowds are depicted by intercutting many different areas of action and representatives of different social classes. But if a filmmaker forswears montage, as Van Sant largely has in Elephant, how can he encompass different points of view? Van Sant has set himself a difficult problem and he crafts an unusual solution, a solution inspired by Tarr’s Satantango: he repeats scenes and photographs them differently each time.
In the most-cited example, the camera precedes shutterbug Elias (Elias McConnell) down the main school corridor. He meets his friend John, asks him to pose for a picture, John does so, they chat briefly and continue on their separate ways, the camera reversing itself and following John down the hall and out of the school. Barely noticeable in the background is a long-legged girl in a red shirt running down the hall past the two boys (we’ve seen her previously on the school athletic field, but probably don't make the connection on the first viewing). Sixteen minutes later, without warning, we are sent back in time and observe this scene again. This time the camera follows Elias into the scene and we view the action from the opposite side of the “action line” or 180-degree axis established previously. Again, the girl scurries by, this time away from the camera in the middle ground of the shot. Since the first version of this scene we have actually met his girl, Michelle (Kristen Hicks). But because of her placement in the frame we still might not notice or identify her. When we view this scene a third time the camera follows Michelle. This time we recognize her and know that she is hurrying from gym class to duties in the school library. She and the camera rush past John and Elias without pausing. Shifting the perspective of the previous scenes, Michelle is now the figure of identification while Elias and John are extras just barely visible and audible in the blurry background.
To resolve this dilemma of point of view, to depict the scene thoroughly, to be “objective” in the journalistic sense, Van Sant violates temporal continuity—one of the qualities associated with long takes. There are cinematic precedents for viewing an action from different points of view. Citizen Kane (1941), Rashomon (1950) and JFK (1991) are among the most famous. Wonderland (2003) and Eternal Sunshine of the Spotless Mind (2004) are two of the most recent. In these films characters relate their experience of a person or event and come to different conclusions, demonstrating the subjectivity of truth.
But Elephant doesn't work that way. The action within each version of this unremarkable scene is scrupulously identical, while the visual differences derive from where the camera is, not the subjective interpretation of each character. Van Sant is addressing more purely cinematic questions. From whose point of view should dramatic action be viewed? Aren't characters in the background just as important as characters in the foreground? Shouldn't what leads up to the scene and what follows it be included as well? When dramatizing Columbine who should the camera follow? The killers? The victims? The survivors? The implication is that if a simple event like two people meeting in a corridor has so many facets then an event like the Columbine shootings may be entirely too complex to depict. The issue of point of view is also embedded in the title of the film, which derives in part from the story of the blind men and the elephant. Each man, according to the fable, has an inadequate understanding of what an elephant is because each one touches a different part of the elephant. They're all right but also wrong. Likewise, each version of the Elias-John-Michelle scene is true, but also incomplete.
Van Sant may want to avoid the “false rhythm” of traditional continuity editing but these repetitions yield an even stranger rhythm. Van Sant is reaching back before Griffith and Eisenstein and The Great Train Robbery to Porter’s truly archaic The Life of an American Fireman (Copyright Version, 1903) in which the rescue of a woman from a burning house is shown in its entirety from two different angles that are edited sequentially rather than intercut (Cook 23). In the temporal design of the film as a whole, Van Sant has actually turned parallel montage inside out by crosscutting two sets of scenes (the day of the shooting and the “Alex & Eric” scenes from the previous day) that cannot be happening simultaneously.
The viewer is placed in a state of almost constant temporal dislocation. No punctuation identifying time shifts is provided, various scenes and pieces of action are repeated, and scenes are held in limbo and continued many scenes later. Nathan and Carrie’s hallway scene is “paused” for 88 minutes and resumes when the shooting commences. The scenes taking place on the day of the shooting describe a sequence of events that actually takes less time than the screen time—a rarity in narrative film. The film, then, is a temporal puzzle picture—Reservoir Dogs (1992), Pulp Fiction (1994), Kansas City (1996), Run Lola Run (1998), Momento (2000), Code Unknown (2000), Amores Perros (2000), Mulholland Drive (2001), Kill Bill (2003-4), 21 Grams (2003), Eternal Sunshine of the Spotless Mind (2004)—in which the viewer must mentally reassemble the scenes to understand their linear chronology.
While Van Sant acknowledges the limitations of the camera he creates even more frustrating omissions through editing. Van Sant (who also edited the film) deliberately truncates scenes so that the final outcome is withheld from us. Before Michelle is shot in the library Elias, who happens to be standing nearby, snaps a photo of Alex. Alex merely glances blankly in Elias’ direction (perhaps he likes the idea of being documented), turns back to Michelle, and starts shooting. Will Elias be one of the victims? Will he live to develop these photos? We never find out (this question has even provoked a debate on Elephant’s Internet Movie Database message board). Later, Alex bursts into the girls’ bathroom and confronts Jordan, Nicole, and Brittany. Will he shoot? Will he find out about the girl hiding in a nearby stall? Unknown.
The film concludes in the school kitchen. Alex has cornered Nathan and Carrie in the freezer (slabs of meat hang behind Alex, underscoring his own butchery). The terrified couple pleads with and rebukes Alex, who recites a childish lyric and pins them down with his weapon. The camera tracks backwards away from the scene. Cut to the closing titles. We never find out if Alex shoots one or both of them. We never find out what finally happens to Alex. We never find out how many people are killed, who they are, or how the community responds to the tragedy. These omissions frustrate some of our basic film viewing assumptions: when a crucial action is set in motion (such as a stand-off involving a gun) we will be shown the conclusion; the film will provide a summary or overview of events; we will find out the fates of the main characters.
One suspects that these omissions occur in part because Van Sant wanted to avoid turning the film into an action film spectacle which would make audiences eager for violence rather than horrified by it. He doesn't want viewers to respond to the carnage the way Alex and Eric respond to their own gunplay: “Whoa, dude, that was awesome.” Van Sant says of the staged violence,
Van Sant doesn't overwhelm the viewer with brutality in the manner of, say, of Darren Aronofsky’s Requiem for a Dream (2000). He practices restraint—a rare quality in mainstream cinema today. The dramatic result of these omissions is to give the climax the feeling of something so swift, unexpected, and chaotic that the film simply can't take it all in, of events so horrifying that the film’s “memory” has repressed them before we have a chance to see them.
Although this withholding of information is at its most extreme during the shooting, throughout the film we are made aware of things happening off screen, as in the football game and the GSA meeting. The first shot of the film (a time-lapse view of a sky changing from dusk to night) is accompanied by the sound of young people playing football (presumably the boys later seen playing on the school athletic field). We hear them but don't see them. Alex and Eric’s entry into the library, where they will begin their rampage, occurs off screen. A medium-close shot of Michelle accompanies the sound of guns being cocked. Michelle turns and glances past the camera with a frown. But the expected point of view shot of what Michelle is looking at is omitted. Instead, the film shifts back to Alex’s bathroom earlier in the day. We must intuit that Alex and Eric entered the library. There is no immediate visual confirmation. Van Sant’s impatience with “the convention of the reaction shot” is evident here.
Many of the adults are only partially viewed, suggesting that adults and teenagers live in the same space but occupy different worlds. In the first dialogue scene, when John takes the wheel of the family car in place of his hung-over father (Timothy Bottoms), the camera pans between the two characters, framing them individually. Photographing two people sitting together in the front seat of a car is a very common type of shot, and Van Sant makes a special effort to keep them separated. They are framed together only briefly (and are visually reunited during the massacre—the only redemptive element in the climax). The scene concludes with father and son talking, but with Mr. Robinson in the center of the shot and John almost completely outside the frame.
Similarly, in the school office Nathan, Carrie, and John talk to administrators who are almost never seen when spoken to. When Michelle walks from the athletic field to the gym the PE teacher who scolds her drifts in and out of a tracking shot that favors Michelle. In the breakfast scene Alex’s mother and father speak but their faces are framed out of the shot. Later, the driver who delivers the gun is off-screen when he speaks to the boys.
This isolation of characters is also effected through selective focus. When John makes a phone call to his home the principal, Mr. Lewis, (Matt Malloy) enters the shot in the background, which is out of focus. But rather then shifting focus back to Lewis when he speaks the focus remains on John in the foreground and Lewis remains a just-recognizable blur (he is additionally obscured by a plexiglass panel that juts into the frame). Not all the tracking shots use depth of focus. In many shots the space beyond the characters is a blur (rather unlike Welles’ and Kubrick’s deep focus tracking shots). The characters often seem to be moving within a sort of bubble (as in the hallway tracking shot following Michelle cited earlier), the world around them glowing but ill-defined.
When Michelle changes clothes in the girl’s locker room the camera views her from above in a medium-close shot. We hear voices muttering derogatory phrases, presumably about her. We can see some girls out-of-focus in the background but can't identify them or be completely certain they are the ones talking. Since Michelle is changing clothes sitting down, and was ordered earlier by her gym teacher to wear shorts instead of sweats to PE, we suspect that she is shy, perhaps embarrassed, about her body. This tight shot effectively conveys her social isolation and desire to hide herself from the world. The off-screen ridicule demonstrates her justified paranoia and the type of bullying which motivates Alex and Eric (Michelle and the killers are linked editorially several times in the film).
These exclusionary techniques often produce a disjunction of cause and effect. At the beginning of the “Alex & Eric” sequence, the camera begins on a science teacher speaking at the front of the class. The camera pans to students asking questions. One male student with a crew cut (one of the “dumb-ass jocks” the shooters despise) turns around and throws something towards the back of the class. The camera pans over to Alex, who is splattered with spitball particles. The standard way to render this scene would be to show the jock throwing, cut to Alex being hit, then cut to reaction shots of the jocks and other students. But Van Sant shoots it so the jock appears to be throwing something at no one in particular and Alex appears with the mess already on him (the behind-the-scenes documentary of the film suggests that this was how the scene was staged). Then, more glop flies in from off screen and pelts Alex. But was it thrown by the crew cut or by the boy in the letterman jacket next to him? (which appears to be Nathan). Thus in one shot we see a character throw something at an unseen target and see another character hit with something thrown by an unseen assailant. Cause and effect, action and reaction are split. Each party—the teacher, the jocks, Alex—are part of the same shot, the same continuous geographic space, but are visually isolated through framing, existing in separate worlds. This suggests disunity, discord, and disconnection that contradict the wholeness—the Bazinian-humanist assumptions—communicated by the long takes and wide-angle views.
These unexpected repetitions, deliberate omissions, and departures from the continuity style of classical Hollywood cinema move the film away from the nominally “realist” long-take tradition of Bazin-Renoir-Welles and closer to the modernist, anti-illusionist theories and practices of Bertholt Brecht and Jean-Luc Godard. Godard has made the denaturalizing of film language—the split of cause and effect—his life’s work. Contempt (1963) concludes with a car crash that kills the central couple. But it occurs off screen, the crucial moment omitted, like the unseen crucial moments during Elephant’s shootout. Godard invented the deliberately disjointed shootout in his first feature film, Breathless (1959). Godard has described his early film practice as “research in the form of spectacle” (Milne 181). Van Sant describes Elephant as investigation in the form of drama:
The relationship of photography (and by extension, cinema and all modes of representation) to “reality” is addressed in a sequence shot in which Elias, walking to school with his still camera, asks a young punk couple if he can take their picture. They consent, but he doesn't “document” them in a neutral or objective manner. He directs them the way a film director would, telling them where to walk, how close to stand together, and directing them to kiss. Photography—once considered a purely mechanical, even inartistic, art—is just as subjective as any other art form. A metaphor for Van Sant’s use of Columbine as subject matter is suggested. Just as Elias “finds” the couple and then manipulates them for his own purposes, Van Sant found a pre-existing subject and shaped it to suit his own interests. The subsequent photography lab and darkroom scenes show how much work actually goes into making a photograph and how the results can be manipulated further in the printing stage. Images don't just “happen,” but are laboriously and consciously crafted.
Elias can be seen as a director surrogate in this and other scenes. Like Van Sant, he casts unknowns and documents outsiders. Elias and the punk couple engage in improvisation, just as Van Sant did with Elephant’s actors. Although Elias uses a still camera, he walks ahead of the couple in the manner of the tracking shots so abundant in Elephant. The punk boy is eager to do a nude scene, but Elias vetoes the idea, just as a film director might veto an over-eager actor’s suggestions. (This indirectly raises the question of who suggested Alex and Eric shower together—the actors or the director?)
The amount of time the film devotes to the process of photography can also be read as a comment—even complement—on Van Sant’s shooting style, which relies less on post-production manipulation of the image than most Hollywood productions today. The lingering shot of Elias dutifully rotating a film-developing canister in his hands as if it were a talisman makes film-based photography a process of almost medieval artisanal patience. Van Sant’s old school impulses are confirmed by the fact that Elephant was edited on film, a method becoming just as rare as the film bath. Just as Elias cuts his processed film with scissors, Van Sant cut his movie with a splicer. In the age of the digital image and the posthuman narrative, Elephant adheres to a 20th century cinematic ideal.