David Bordwell's website on cinema   click for CV

Home

Blog

Books

Reinventing Hollywood: How 1940s Filmmakers Changed Movie Storytelling

Film Art: An Introduction

Christopher Nolan: A Labyrinth of Linkages pdf online

Pandora’s Digital Box: Films, Files, and the Future of Movies pdf online

Planet Hong Kong, second edition pdf online

The Way Hollywood Tells It pdf online

Poetics of Cinema pdf online

Figures Traced In Light

Ozu and the Poetics of Cinema pdf online

Exporting Entertainment: America in the World Film Market 1907–1934 pdf online

Video

Hou Hsiao-hsien: A new video lecture!

CinemaScope: The Modern Miracle You See Without Glasses

How Motion Pictures Became the Movies

Constructive editing in Pickpocket: A video essay

Essays

A Celestial Cinémathèque? or, Film Archives and Me: A Semi-Personal History

Shklovsky and His “Monument to a Scientific Error”

Murder Culture: Adventures in 1940s Suspense

The Viewer’s Share: Models of Mind in Explaining Film

Common Sense + Film Theory = Common-Sense Film Theory?

Mad Detective: Doubling Down

The Classical Hollywood Cinema Twenty-Five Years Along

Nordisk and the Tableau Aesthetic

William Cameron Menzies: One Forceful, Impressive Idea

Another Shaw Production: Anamorphic Adventures in Hong Kong

Paolo Gioli’s Vertical Cinema

(Re)Discovering Charles Dekeukeleire

Doing Film History

The Hook: Scene Transitions in Classical Cinema

Anatomy of the Action Picture

Hearing Voices

Preface, Croatian edition, On the History of Film Style

Slavoj Žižek: Say Anything

Film and the Historical Return

Studying Cinema

Articles

Book Reports

Observations on film art

Archive for the 'National cinemas: UK' Category

One last big job: How heist movies tell their stories

The Underneath (1995).

DB here:

The announcement of Ocean’s Eight (premiering, when else?, on 8 June next year) reminded me of the staying power of the heist genre, also known as the big-caper film. I discuss it a bit in Reinventing Hollywood, my book on 1940s storytelling, but it developed and spread out most vigorously from the 1950s on.

In its origins it relies on masculine roles; if women are present they’re likely to be at best helpers, at worst traitors. The big job is likely to be endangered by a man telling his wife or mistress too much, and then the police or rival gangs may interfere. So the news that Ocean’s Eight will center on a female gang constitutes an intriguing wrinkle. (Will a boyfriend try to spoil the caper?)   Warners’ ambitions for another trilogy seem evident: presumably the entry is Eight because Warners hopes for a Nine and a Ten. Commenters are already worried that a “feminized” version runs the risk of flopping the way Ghostbusters did last year, but I’m more hopeful. The reason is that clever storytelling is encouraged in this genre. There’s a chance that the film, produced though not directed by Steven Soderbergh, will show us that this old dog has new tricks.

Because I’m interested in how the storytelling strategies of popular cinema, the heist film is a natural thing for me to consider. Refreshing the genre may involve not just adjusting the story world—giving men’s roles to women—but also considering ways of handling two other dimensions of narrative: plot structure and cinematic narration. I argue in Reinventing that Hollywood filmmaking uses a sort of variorum principle, a pressure to explore as many narrative devices as possible within the constraints of tradition. For this reason, the prospect of Ocean’s Eight prodded me to think about how convention and innovation work in the caper movie. It’s also a good excuse to go back and watch some skilful cinema.

 

From the fringes to the core

The Asphalt Jungle (1950).

It’s useful to think of a genre as a category having a core and a periphery. At the core are prime, “pure” instances: prototypes. Stretching out from it are less central, fuzzier cases. Prototypical musicals are Footlight Parade, Shall We Dance?, Meet Me in St. Louis, and My Fair Lady. Somewhat less central are concert films and musical biopics like Lady Sings the Blues and Coal Miner’s Daughter. At the periphery are movies with a single song or dance number. Of course the category changes across history; a critic in 1960 might have picked North by Northwest as a prototypical spy thriller, but ten years later a Bond film would have probably held that place.

The prototypical heist film, critics seem to agree, doesn’t just include a robbery. There are robberies in films about Robin Hood and gentleman thieves like Raffles and the Saint. The outlaw or bandit film like They Live by Night and Bonnie and Clyde contains a string of robberies, but they wouldn’t be at the center of the heist category.

Donald Westlake proposed a concise characterization: “We follow the crooks before, during, and after a crime, usually a robbery.” This indicates two things. First, the plot is structured around the big caper, though there might be lesser crimes enabling it, such as stealing weapons. Second, the viewpoint is organized around the criminals, not the detectives who might pursue them, as in a police procedural. By these criteria, High Sierra (1941), although allied to the bandit tradition, fits because most of the film concerns a single robbery, its preparations and aftermath.

Westlake shrewdly goes on to note that the interest of a heist plot inverts that of the traditional locked-room detective story. Instead of wondering how something difficult was done, we wonder how something difficult will be done. “The puzzle exists before the crime is committed.” To fill in the how, crooks with specific skills (safecracking, demolition, getaway driving, and so on) are recruited. Because the process matters so much, Westlake adds that heist plots focus on details of time and physical circumstance, and they draw attention to impediments such as “dogs, locks, alarms, watchmen, and complicated traps.” Most of these features are missing from High Sierra; we’re not encouraged to speculate how the job will be pulled, there’s little meticulous planning and no division of labor by expertise, and the only dog is friendly. So it’s unlikely to be a core instance of the genre as we now consider it.

Most critics consider The Asphalt Jungle (novel, 1949; film, 1950) a solid prototype. In fiction and film it’s not the very earliest instance, but it displays what we might call the “completed arc” of a heist plot. A gang of varying talents is assembled and funded to pull off a jewel robbery, which succeeds partly at first and eventually fails.

A less famous example comes from 1934, Don Tracy’s novel Criss-Cross. Here an alienated loner finds that his long-ago girlfriend has married a local crook. To get her back, the hero agrees to be the inside man in a robbery of the armored truck he drives. The heist goes badly, with the hero shooting the gangster and getting shot himself. He’s acclaimed as a hero, but fate catches up with him when a blackmailer who knows the truth starts putting the squeeze on him. Where The Asphalt Jungle ranges across many characters’ knowledge in a fairly tight time span, Criss Cross restricts its point of view to the protagonist but traces his life over a long period. These storytelling options, ingredient to any narrative whatsoever, will shape how the heist plot is handled.

Criss-Cross and The Asphalt Jungle show that we’re dealing with an exceptionally schematic plot pattern. For convenience we can distinguish five parts.

Circumstances lead one or more characters to decide to execute a heist (robbery, hijacking, kidnapping).

The initiators recruit participants.

As a group they are briefed and prepare their plan. They study their target, rehearse their scheme, or take steps to make it easier.

The heist itself begins and concludes.

The aftermath of the heist, failed or successful, shows the fates of the participants.

There are other conventions, nicely laid out by Stuart Kaminsky in 1974 and more recently by Daryl Lee. But I’ll just concentrate on this structure and how it governs narration, because these aspects show very clearly the constant dynamic of schema and revision in the filmmaking tradition.

 

Four capers, mostly unhappy

Criss Cross (1948).

Four films in the period covered in my book point toward the heist genre, though only one became a core instance. The presence of a “consolidating” film might be one way genres emerge.

The earliest candidate is an unlikely one. In 1941 Laura and S. J. Perelman mounted a notably unsuccessful play called The Night Before Christmas, a farce in which crooks buy a luggage store in order to tunnel from its basement into the bank vault next door. Although Paramount was a backer of the stage show, Warners bought the rights and revamped it for Edward G. Robinson, who had had success in gangster comedies like A Slight Case of Murder (1938). The adaptation, Larceny, Inc. (1942) maintained the premise but revised the plot, multiplying the complications that face the leader Pressure and his two minions. The twist is that thanks to Pressure’s daughter, who’s unaware of the scheme, business starts booming and the crooks find themselves prosperous businessmen. Woody Allen swiped the premise for his Small-Time Crooks (2000).

Just as we don’t expect a comedy to be the initial entry in a genre mostly associated with drama, we might expect that the early entries in the genre would be the tidiest, with complex variations building on them. Actually, two precursors of The Asphalt Jungle display remarkable complexity.

The simpler one, Criss Cross (1948), starts late in the preparation phase: the night before the caper, when the gang is holding a party to distract a local cop who has his eye on them. We learn that Steve is working with Slim’s wife Anna to double-cross the gang. The next day, Steve is driving the armored car and recalling how he got into this situation. A flashback provides the circumstances for the robbery—Steve’s pursuit of Anna, her marrying Slim, and Steve’s impulsive proposal of a robbery to protect Anna from Slim’s punishment.

We return to the armored car’s routine briefly before we flash back to the planning phase. The robbers assemble and the task is reviewed by a mastermind they hire. This second flashback ends and, back in the present, we see the heist itself. The aftermath of the bloody robbery shows Steve in the hospital, honored by the community. A thug kidnaps him and takes him to Anna. Slim has survived the gun battle and finds the couple; he kills them before he is mowed down by the police.

Early as Criss Cross is in the history of the film genre, the plot components are already being scrambled. Thanks to 1940s Hollywood’s commitment to flashbacks and character subjectivity, the novel’s linear action gets fractured, compressed, and stretched. Moreover, in most films, the decision to pull the heist is made fairly early. Here and in the novel, circumstances slowly force Steve to rescue Anna from the man who mistreats her. The buildup to this decision consumes nearly forty minutes of an 87-minute movie. Similarly, the aftermath of the heist is quite protracted, running about twenty minutes. And whereas other films (and novels) dwell on the recruitment of thieving talent, here this is elided. That’s because the other members of the gang are less significant than the central triangle of Steve, Anna, and Slim.

The Killers (1946) is even more complex because it clamps the block construction of Citizen Kane onto the heist schema. We start with one bit of the heist aftermath: the death of the gas-station attendant Ole. Like Kane, who also dies in his bed at the start of his film, Ole is the protagonist of the past-tense story. We next meet the protagonist of the present-time story, the insurance investigator Riordan, who functions like the reporter Thompson in Kane, and sometimes even gets the same silhouette treatment.

          

The rest of the film splits up the basic parts of the heist schema, rearranges their order, and presents them through a series of several flashback blocks, anchored in six characters’ testimony.

The heist itself is very brief, running only about two minutes, and the planning phase is only a little longer. As in Criss Cross, the leadup and the aftermath constitute the bulk of the film, but these phases are broken up into many bits, and these are often shuffled out of chronological order. We see Ole’s discovery by Big Jim years after the heist and then we see Ole, days after the heist, distressed that his lover Kitty has betrayed him. Similarly, a crucial event after the heist—Ole’s stealing the loot from the gang—precedes a scene before the heist in which he learns the gang is planning to double-cross him. We have to reconstruct the canonical sequence, from circumstance to aftermath, out of fragments revealed in testimony taken by Riordan.

Critics haven’t taken Criss Cross and The Killers as the first core instances of the heist film, perhaps because these movies complicate the classic plot template. In addition, they downplay the planning and execution phases. The heist scenes are relatively minor compared to the time devoted to other phases, particularly the circumstantial buildup. Moreover, the robber teams aren’t vividly characterized and don’t display a range of expertise. The emphasis falls on Steve and Ole, both played by Burt Lancaster, as the beautiful losers who become suckers for treacherous women and their brutal men.

By contrast, the linearity of The Asphalt Jungle throws the plot template into crisp relief. Both novel and film offer a full-blown enactment of central features picked out by Stuart Kaminsky. The gang, despite low social standing, has many skills. They’re held together by a leader who may also be a brainy planner. The robbers’ plan requires “skill, practice, training, and above all, perfect timing.” The gang’s resourcefulness is tested by accidents. The caper is partly successful but largely a failure, often through sexual weakness on the part of some players.

W.R. Burnett’s book opens up the wide narrative possibilities of the genre, thanks to an expanded group of characters. Where Don Tracy’s novel Criss Cross confined itself to relatively few figures, Burnett innovates by developing no fewer than eighteen distinct individuals, all somehow connected to the jewel robbery. New roles are assigned that will be staples of the genre—financier, go-between, bent cop, owner of a safe house. We see policemen of all ranks, a news reporter, the master mind Dr. Riedenschneider, the fixer Cobby, the bankroller Emmerich (who has a wife, a mistress, and a hired detective), and the four men on the team, two of whom have women attached.

The novel lays out the canonical action phases with surgical efficiency. But as in the films I’ve just mentioned, the planning and the heist itself take relatively little space. Given so many characters’ contribution to the action, their gradual convergence and their dispersed fates dominate the book. Over a hundred pages are devoted to setting up the circumstances and recruiting the gang. The aftermath, tracing the outcome for all involved, runs another 140 pages.

The film adaptation retains the novel’s scope, trimming only a couple of minor characters. It moves briskly among the ensemble, attaching us to several in turn, so we know more than any one character does. The narration creates parallels among the team members and their affiliates (such as four women of different classes, with varying commitments to their men).

The result is something of a network narrative, a cross-section of several people whose lives change because of the robbery. There’s no clear-cut protagonist like Steve in Criss Cross. Dr. Riedenschneider has the most screen time, but the film begins and ends with the hooligan Dix, and the fate of both men dominates the climax. I point out in Reinventing Hollywood that a multiple-protagonist film often gives extra weight to one or two characters.

 

50s and 60s stylings

Violent Saturday (1955).

With The Asphalt Jungle as a robust prototype, writers and filmmakers developed the heist plot throughout the 1950s and 1960s. Despite his clumsy style (”He nodded with his head”), Lionel White became known as “the king of the caper novel” for his tireless variants on robberies, hijackings, and kidnappings. The other major figure was the far superior Donald Westlake, who  wrote brutal heist tales as “Richard Stark” and comic ones under his own name.

In the comic register, Westlake was beaten to the post by heist films from England. The Lavender Hill Mob (1951) showed a timid clerk engineering the theft of gold bullion, which he disguises as kitschy miniature Eiffel Towers. The Lady Killers (1955) centered on a gang fooling their landlady into thinking that their planning meetings are actually string-quartet sessions. In Italy, the classic Big Deal on Madonna Street (I soliti ignoti, 1958) showed five incompetents trying to break into a pawnshop safe.

Big Deal was considered a parody of a much more serious film, one that was by then a landmark: Rififi (Du rififi chez les hommes, 1955). It did a great deal to consolidate the genre in France, along with Touchez pas au grisbi (1954) and Bob le flambeur (1956). (Both The Killers and Criss Cross had been imported in the Forties.) In England, the heist drama was seen in The Good Die Young (1954), while in the US there were 5 Against the House (1955), Violent Saturday (1955), Plunder Road (1957), The Big Caper (1957, from a White novel), and Odds Against Tomorrow (1959). The year 1960 saw a remarkable number of releases, with The League of Gentlemen (U.K.), The Day They Robbed the Bank of England, Seven Thieves, and most famously Ocean’s 11 (all U.S.).

Through the 1960s the genre proliferated, including a remake of The Asphalt Jungle (Cairo, 1963), Jules Dassin’s near-self-parody of Rififi (Topkapi, 1964), and a string of heist films that mixed in romantic comedy (e.g., Gambit, 1966; How to Steal a Million, 1966). The Italian Job (1964) became a cult movie, with tourists visiting the sites of shooting in Turin. The genre has remained robust throughout world cinema ever since.

The very rigidity of its format and the constraints it lays down, have allowed for ingenious innovations. I’ll sample some of those from the genre’s first dozen or so years, concentrating on three areas: the differing emphasis given to the plot phases; variants of point-of-view; and manipulations of time.

 

Where’s the action?

Rififi (1955).

The Asphalt Jungle offers a fairly pure geometry: the heist, lasting about eleven minutes, ends just before the film’s midpoint. On the other side of that is another big scene: a gunfight initiated by Emmerich’s henchman Brannom when Dr. Riedenschneider and Dix bring in the loot. That confrontation takes another seven minutes, so the heist and its immediate wrapup sit at the center of the running time. What led up to the heist, the circumstances, recruitment, and planning phases, consume the first forty-five minutes of the film, and the aftermath of the heist, affecting the surviving characters, will take just about as long. This balanced geometry, putting the heist snugly at the center of the plot but devoting most time to showing many lives leading up to and away from the crime, suits the extensive characterization of all involved.

This is a neat structure, but not the only possibility. Any phase of the action can be given great weight. The first half hour of Bob le flambeur is an episodic introduction to Bob’s routines and associates. It’s the circumstantial phase—he’s running out of money—presented at low pressure, emphasizing instead the fascinating milieu Bob coolly drifts through. Only when Roger points out a croupier at the Nantes casino does Bob formulate “the job of a lifetime.” Then crew assembly and planning can begin. As the plot develops, characters and relationships presented casually in the exposition become crucial to the heist and its unraveling.

Other phases can be stretched out. Ocean’s 11 dwells on the phase of gathering the talent. Before the start, Danny Ocean has already conceived the heist, but we aren’t told the plan. Instead we watch his team members converge on Las Vegas. Not until the film’s midpoint, at 53:00, is the plan revealed in a briefing. The dawdling exposition suits a film about cool guys hanging out.

The formation of the crew is exceptionally protracted in Odds Against Tomorrow, because here two members hesitate about participating. The racist Earl Slater, wracked by the shame of not providing for his wife, finally gets her to agree to let him join. Not until twenty-three minutes from the end does the African-American musician Johnny Ingram finally accept a role in the heist, which immediately becomes the film’s climax.

The source for Odds Against Tomorrow (one of the better heist novels) handles the basic pattern very differently, putting the heist just before the midpoint and devoting the second half of the book to the increasingly hopeless getaway, where the interracial tensions play out at length.

The planning phase sometimes includes a rehearsal for the heist, and that may require a subsidiary crime, such as stealing keys or a vehicle. The League of Gentlemen has an unusually long planning phase, in which the gang masquerades as soldiers in order to seize weapons from a military post. This robbery, running nearly twenty minutes, takes longer than the film’s main heist but serves to demonstrate the team’s resourcefulness and esprit de corps. The men’s precision and resourcefulness (based on their wartime experience) lead us to expect a successful main event.

Kaminsky suggests that the big-caper genre resembles the soldiers-on-a-mission movie, which centers on the details of a process executed by skillful specialists. That display of process comes to the fore in The Asphalt Jungle’s central robbery, which initiated the convention of detailing the mechanics of the heist. Other films expanded the heist sequence to larger proportions than was evident in The Killers or Criss Cross. Rififi became the model; its heist runs twenty-five minutes without any words being spoken. (That sequence is part of an even longer wordless stretch that runs over thirty minutes.) Dassin’s film meticulously breaks down the steps of a complicated robbery, allowing us to appreciate the men’s clever planning while, as usual, ratcheting up moments of suspense when it appears they might fail.

After Rififi, heist scenes got longer and more intricate. The one in Big Deal on Madonna Street runs twenty minutes, those in Seven Thieves and The Day They Robbed the Bank of England run thirty minutes, and those in Topkapi and The Italian Job approach forty minutes. To motivate these long, virtuosic passages, there need to be many obstacles, many ingenious ways around them, and many characters concentrating on their assigned duties.

Long or short, the heist can be shifted off-center. The heist can be the climax, as in The Big Caper, or it can be virtually the film’s entire second half, as in The Italian Job (which arguably doesn’t complete the heist and so never provides a proper aftermath). Topkapi’s flamboyant robbery, involving elaborate subterfuges, dominates the film’s latter stretches, leading to a quick reversal in its epilogue. The heist is finished early on in The Lady Killers, since the crucial action is the long aftermath in which, contradicting the title, the crooks dispose of one another. By contrast, the heist dominates nearly all of Larceny, Inc. Most of the film is devoted to the crooks’ slowly deepening tunnel, as the process is interrupted by breaks in the water main, a geyser of furnace oil, and their decision to abandon their plan in favor of going straight—before another crook comes along to continue the caper.

Or the heist can launch the film. In Plunder Road, after the train robbery is completed, the rest of the film is the aftermath, consisting of chases and suspense sequences. The circumstances, recruitment, and planning phases are alluded to in dialogue, as suits a low-budget production. The plot of Touchez pas au grisbi starts the morning after the caper and shows thieves trying to protect their loot from a treacherous drug dealer. It was probably inevitable that somebody would make a heist movie that keeps the heist offscreen and makes the action one long aftermath.

 

Steering the moving spotlight

Topkapi (1964).

Many films based on mystery rely on restricting a character’s range of knowledge, as in detective novels told in the first person. But suspense, Hitchcock and others recognized, relies on greater access to story information. If there’s a bomb under the table, Hitch advised, tell the viewer but don’t tell the characters. The heist film, as an account of how a crime is committed, depends more on suspense than mystery. So we’d expect a constant shuttling from character to character, filling us in on how the scheme is going and letting us worry about impending dangers.

This is generally what we find. Criss Cross is unusual in confining us to what Steve knows about Anna’s and Slim’s real agendas, and this confinement is what yields the surprises that pop up in the film’s final stretches. Mostly, though, heist movies and novels use omniscient narration. Employing what I call in Reinventing Hollywood the “moving spotlight” approach, the film shifts us from attachment to one character to another.

The moving spotlight permits us to understand the choreography of the crime, in which many characters play a part, and to feel suspense when matters of timing or accident intrude. Indeed, the convention of the unforeseen accident that spoils the heist relies on our breaking attachment to the crooks and getting privileged access to the off-schedule night watchman, the passing witness, or the failed piece of machinery. In Topkapi we and not the thieves learn how a stray bird will upset their getaway.

This omniscient narration (which need not tell us absolutely everything) can be deployed in a range of ways. One option is chapter titles, as in Big Deal on Madonna Street (and revived by Tarantino in Reservoir Dogs). A more common resource is the non-character narrator, present as an authoritative voice-over. In Bob le flambeur, a worldly male voice introduces us to Bob’s daily routine and comments somewhat cynically on his place in his milieu.

A similar all-knowing voice introduces us one by one to the heisters in The Good Die Young; there it eases the transition into expository flashbacks. Topkapi treats this convention, as others, with self-conscious playfulness. At the beginning the beguiling woman in the gang addresses the camera and explains that the heist aims to get the emerald she covets. She doesn’t take this narrating role again until the very end, in a brief epilogue that replays the gathering of the gang in stylized form.

Heist films must often choose whether to widen the field of view to include either the cops or civilians. Armored Car Robbery (1950), for me a peripheral instance of a heist film, alternates the gang’s activities—which do pass through the phases I’ve outlined—with the detectives’ investigation. The result is balanced between heist film and police procedural. Something similar happens in Violent Saturday (1955), which devotes most of its plot to a gang’s scheme to rob a mining town’s bank. Crosscut with the gang’s gathering and planning are scenes showing several members of the community, each with domestic problems. These scenes, typical of small-town melodrama (dissolute rich people, sexy nurse, repressed librarian, lovelorn bank manager), not only pad out the film but show us characters who will be affected by the heist. Some become witnesses and bystanders during the robbery, one is killed, and one, a local mining manager, will be forced to fight the gang at the climax.

The timing and selection of the shifts among characters can build up specific effects. For instance, in Big Deal on Madonna Street, the heist is conceived by Cosimo, a hardheaded crook doing time. A handful of would-be crooks decides to execute the robbery without him. We know, as the gang does not, that he has been released and is conspiring to interfere. The filmmakers could have made Cosimo’s return a surprise revelation, but instead our expectation that he’ll start meddling intensifies our interest.

Similarly, we know, as the Topkapi gang does not, that Arthur is a police spy. But then he confesses to them, so we know, as the Turkish police do not, that he is to be a double agent. The film milks suspense during the scenes in which Arthur relays information to the authorities, and then it makes us wonder how the gang will use their new knowledge of police surveillance. This is the sort of fine-tuned choice about “who knows what when” that heist films must make at every turn.

At a more granular level, when the spotlight is on one character, the filmmakers must choose whether to make those moments, shot by shot, highly restricted or not. In The Killers, Nick Adams’ account of Ole’s reaction to the stranger stopping for gas is more or less plausibly limited to Nick’s ken. But the precise exchange of glances between the driver and the Swede signal to us that Ole has been recognized. Nick doesn’t register this, since he’s standing at the rear of the car (barely visible in my first still, over Ole’s right shoulder).

          

In effect, this is a narrational aside to the audience. Nick can’t see this significant exchange of glances, and his voice-over doesn’t specify it, so we can’t assume Riordan learns of it. But this privileged moment primes us to see Big Jim as a major menace in scenes to come. This sort of deviation from what a character could plausibly know or notice is common in cinematic narration and differentiates it from literary narration, which is more tightly restricted in handling “point of view.” Classical Hollywood is biased toward unrestricted narration; radical restriction is rare. The heist genre can exploit many fine gradations between attachment to a character and more wide-ranging knowledge.

 

What’s our clock?

The Killing (1956).

As we’ve seen, two early instances of heist films break with a linear layout of the story. Criss Cross presents what Reinventing Hollywood calls a crisis structure, beginning just before the turning point and flashing back to show what led up to it. The Killers has a refracted narration, with all of the past events passed to the investigator Riordan and us through witnesses’ testimony. The flashbacks in Criss Cross are chronological, but those in The Killers are drastically out of order.

Time-scrambling, then, seems to be especially welcome in the heist film (as not, say, in the Western). The basic plot pattern is such a simple one that shuffling phases or parts of phases doesn’t create confusion. Yet few films of the 1950s I’ve found display the complexity of the Forties examples. Indeed, a time-juggling heist novel, The Lions at the Kill (Simon Kent, 1959) was ironed into straightforward linearity in the film version, Seven Thieves.

The Good Die Young displays a more complicated structure. It begins at a point of crisis, with four men driving in a car to the site of the caper.

The voice-over narrator introduces each man in turn, and then a long flashback explains his backstory. Deprived of information about the heist, we have to wonder how the narration will bring the men together and create their plan.

After the long section explaining each man’s circumstances, all involving women and money, the flashbacks meld. The men gradually come to know each other through frequenting the same pub. Their desperation pushes them toward accepting one man’s proposal that they rob a post office. With so much time spent on preparation and assembly, the actual planning is necessarily brief. Eventually the final flashback flows seamlessly into the present-time situation, initiating a brief replay of the dialogue in the car that opened the film. The heist (devolving into a disastrous gun battle) initiates the climax, and an airport aftermath concludes the film.

The most ambitious 1950s recasting of the nonlinear efforts of the 1940s films is The Killing (1956). The source novel, Lionel White’s Clean Break (1955) consists of interwoven threads. A crook, his pal, a cashier, a bartender, and a crooked cop conspire to rob a racetrack. They will employ helpers to start a fistfight and shoot a racehorse. The early chapters devote short stretches to following each one, already recruited, through the circumstances phase of the heist template. The men assemble at their planning meeting, which is invaded by the snoopy wife of the weak cashier. She then tells her boyfriend about the heist, so another strand involving other characters is introduced.

On the day of the robbery, White’s narration again attaches itself seriatim to each man as he executes his role in the scheme. While the early part of the book had a few mild jumps back in time to follow each heister, the day of the job creates extreme back-and-forth shifts. We follow the bartender, for instance, going to the track on the fateful day, before the next subchapter skips back to the cashier waking up and then going to the same train. In a later section, we skip back to the previous day and attach ourselves to the sniper who will shoot the racehorse. Because of the overlapping lines of action, the shooting of the horse and the starting of the bar fight are presented twice. Everything eventually converges on Johnny Clay, the mastermind, who breaks into the money room and steals the day’s revenue.

These temporal overlaps emerge partly from the description of the action, but they’re also marked by either characters looking at clocks or watches, or by explicit mention in the prose, such as “It was exactly six forty-five when…”

The film version makes these overlapping schedules much more explicit. A detached voice-over describing police routine or criminal behavior had become a commonplace in the 1940s, in both “semidocumentary” films and in radio drama, notably Dragnet. In The Killing, apart from lending an aura of authenticity, the voice-over exaggerates the time markers of the novel by introducing scenes crisply. The first five scenes are introduced with these tags:

“At exactly 3:45…”

“About an hour earlier…”

“At seven PM that same day…”

“A half an hour earlier…”

“At 7:15 that same night…”

These lead-ins remind us of the ticking clock, set up parallels among the characters, and get us acclimated to the film’s method of tracking one character, then jumping back in time to track another. This is moving-spotlight narration on markedly parallel tracks. The same method will be applied during the heist, when ten consecutive scenes and several others will be tagged in the same way. (“Mike O’Reilly was ready at 11:15.”)

It’s not just the repetitions that exaggerate White’s jagged time scheme. When the sniper Nikki is shot after plugging the racehorse, the narrator reports drily: “Nikki was dead at 4:24.” Cut to Johnny leaving a luggage store, and the voice-over announces: “At 2:15 that afternoon Johnny Clay was still in the city.”

     

     

The time jumps more or less buried in White’s prose are made dissonant in the film. Here the juxtaposition heightens the likelihood that Johnny’s robbery won’t go according to plan.

The replays are no less sharply profiled. As the scenic blocks move from character to character and skip back in time, they include actions we’ve already seen. The most persistent is the repetition of the track announcer’s calling the start of the crucial seventh race, but we also get replays of the wrestler Maurice starting a fight, the downing of the horse, and glimpses of Johnny waiting to slip into the cash room.

Instead of shuffling flashbacks in the manner of The Killers, The Killing offers brief time chunks stacked in slightly overlapping array until they all square up in a single moment, the consummation of the robbery. This structure allows for the sort of character delineation we find in The Asphalt Jungle while also offering the audience a formal game to enjoy. Kubrick’s film can be seen as pointing in two directions—revising the flashback format of the 1940s entries, but becoming a point of reference for later filmmakers eager to innovate games with time and viewpoint that would remain comprehensible to the audience.

The Killing helped make Clean Break White’s most famous novel. It was often republished under the film’s title. Perhaps in a grateful spirit White’s 1960 novel Steal Big (1960) included this scene.

Donovan didn’t look at the half-dozen worn, barely legible signs in the dingy lobby of the building. He went at once to the elevator and asked for the fifth floor. Getting out of the elevator, he turned left, took a dozen steps and knocked on a pebbled-glass door. The door bore the legend, KUBRIC NOVELTY COMPANY.

Mr. Kubric, it turns out, supplies illegal guns and explosives.

 

Art for artifice’s sake

Inside Man (2006).

White’s in-joke, along with The Killing’s gamelike approach to structure, reminds us that whatever its claim to realism, this is a highly artificial genre. The strict template and the ritualistic steps in it—recruiting the crew, casing the target, checking watches—invite filmmakers to tinker with self-conscious narration that lets the viewer in on the joke.

One convention, that of the rehearsal for the big job, is flaunted in Bob le flambeur, when the narrator simply breaks into the film with the line, “Here’s how Bob pictured the heist,” and we get a stylized, hypothetical enactment of the job carried off perfectly. In these films, when the rehearsal goes well, you know the real thing will face problems.

     

Something similar, though suited to light comedy, takes place in Gambit. Here we think we’re seeing the heist as executed (across twenty-three minutes), but it proves to be no more than the fantasy of the crook hatching it. Again, everything that succeeds in the fantasy goes wrong in reality.

As it gained a profile, the genre got reflexive. In the novel that was the source for The League of Gentlemen, the heisters explicitly model their plan on the one in Clean Break. The film version likewise sets its heisters mimicking a pulp novel, though it didn’t specify the title. And in The Wrong Arm of the Law (1963), Peter Sellers’ cockney mastermind promises to show his henchmen “educational films and training films” like Rififi, The Day They Robbed the Bank of England, and The League of Gentlemen.

The opportunities the genre provides for experimenting with story stratagems, sometimes in very self-conscious ways, seem to have been part of the reason later filmmakers tried their hand at it. In the 1990s and 2000s, writers and directors drawn to neo-noir and narrative experiment took up thriller conventions generally, and the heist film was one option.

In Reservoir Dogs (1992), Tarantino revived the shuffling of time and viewpoint that was ingredient to the genre. Although the film pays homage to Clean Break, its form is no less indebted to Criss Cross. Like that film, it flashes back from the day of the heist in order to run through the canonical phases of action. And as in The Killers, those phases are scrambled out of order. Brian Singer’s The Usual Suspects (1995) brings the (very ‘40s) device of the lying flashback to the genre, within the template of police questioning after the heist. Inside Man (2006) mixed to-camera narration, flashbacks, and flashforwards to interrogations after the crime.

Among Americans it’s Steven Soderbergh who has returned to the caper film most frequently. The Underneath (1995) is a remake of Criss Cross, with an extra layer of flashbacks. Logan Lucky (2017), with its deadpan redneck losers, joins the tradition of comic heist movies. Most strikingly, the Ocean’s series makes a virtual fetish of the male camaraderie and playful plot tricks typical of the genre. The films pepper the action with voice-overs, cunning ellipses, and flashbacks within flashbacks. The plots hide key information about the plan. They fill the action with in-jokes, such as a star cameo by Bruce Willis commenting on box-office grosses. Like 1960s films, they incorporate romantic comedy. In Ocean’s Eleven ((2001) Danny and Tess arrange a post-divorce reconciliation, and in Ocean’s Twelve (2004) Danny’s pal Rusty gets involved with police agent Isabel.

Piling up obstacles, reversals, bluffs, and double-bluffs, the films form a kind of anthology of the genre’s tricks. By the time we get to Ocean’s Thirteen (2007), the early phases of the standard plot schema can be given short shrift. We know the gang and its modus operandi, so the bulk of the film becomes a mind-bogglingly intricate heist including everything from planting bedbugs in a hotel room to manufacturing loaded dice in a Mexican factory under threat of strike. The network of rules and roles laid out in the 1950s—master mind, aged expert, financier, crooked helpers, allies and rivals and go-betweens and stooges—are given baroque elaboration and treated with an almost self-congratulatory panache.

 

So am I looking forward to Ocean’s Eight? It’s directed by Gary Ross, so I don’t expect Soderbergh’s casual sheen. And the target of the heist, a fashion show at the Met, may be a bit too on the nose; why can’t the ladies tackle a payroll or a munitions factory? Still, like a butterfly collector looking for new specimens, I’m quite curious. When it comes to narrative strategies, I’m a sucker for a fresh con.


Thanks to Jim Healy and Geoff Gardner for discussion of the genre.

Stuart Kaminsky’s American Film Genres: Approaches to a Critical Theory of Popular Film (Plaum, 1974) is a trailblazing study, and the chapter on the “big-caper” film is an indispensible starting point for studying this kind of movie. Later editions of the book eliminated this chapter. A wide-ranging recent survey is Daryl Lee’s The Heist Film: Stealing with Style (Wallflower, 2014). Alastair Phillips offers an in-depth analysis of a classic in Rififi (Tauris, 2009).

My quotations from Donald Westlake come from Murderous Schemes: An Anthology of Classic Detective Stories (Oxford University Press, 1996), pp. 1 and 126. I also found the entry on Lionel White in Brian Ritt’s Paperback Confidential: Crime Writers of the Paperback Era (Stark House, 2013) helpful. The “KUBRIC” passage in White’s Steal Big (Fawcett, 1960) is on p. 47. White much influenced Westlake, as I hope to show in a future entry, but Westlake was a far superior stylist, as I discuss a little in this entry. See also this note on Levi Stahl’s anthology of Westlake nonfiction.

For more on analyzing narrative see “Three Dimensions of Film Narrative” and “Understanding Film Narrative: The Trailer.” Tarantino’s debts to the 1940s are reviewed in this entry. See as well my recent entry on thrillers. I analyze the convoluted narrative of The Killers in Chapter 9 of Narration in the Fiction Film. The Way Hollywood Tells It: Story and Style in Modern Movies discusses 1990s filmmakers’ relation to the Forties. Multiple-protagonist plotting is considered in Chapter 3 of Reinventing Hollywood: How 1940s Filmmakers Changed Movie Storytelling.

The concept of schema and revision is considered here as applied to style; this entry, like Reinventing, applies it to narrative principles. My Dunkirk entry points out affinities between the overlapped time frames in The Killing and the three-track scheme Nolan gives us. Unlike Kubrick, Dunkirk applies the principle to the entire film and assigns each timeline a different duration, but it generates a back-and-fill organization and clusters of replays reminiscent of The Killing.

Burt Lancaster is considered here, and why not?

Ocean’s Eight (2018).

Wisconsin Film Festival: Cutting to the chase, and away from it

Jeanne 600

Nocturama (2016).

DB here:

Despite my recent jab at D. W. Griffith, I gladly give him credit for making crosscutting a central technique of narrative cinema. Using editing to switch our attention from one story line to another is a fundamental resource of moviemaking everywhere.

Crosscutting is most apparent in those passages of quickly alternating shots that build tension during chases and last-minute rescues. That’s a prototype of what we credit Griffith with consolidating. But crosscutting is used outside such climactic stretches. Hollywood silent features often crosscut story lines throughout the film, without pressure of a deadline and without much happening in some lines of action. It seems to be a way that filmmakers found to keep the audience aware of many story strands.

Crosscutting is a cinematic version of a very old narrative strategy, that of alternating presentation. Once you have several story lines, you can switch among them. Homer does this in the Odyssey, interweaving Ulysses’ wanderings, Telemachus’ efforts to find him, and Penelope’s holding off the suitors.

Homer initially handles these lines in large blocks, in separate “books.” After attaching us to Telemachus in Books 1-4, Homer shifts us to Ulysses for a long stretch. Such interlacing can be found in medieval narrative too, and of course it dominates modern novels, with chapters shifting among action lines and character viewpoints.

Crosscutting large chunks can give way to shorter bursts. Ulysses’ travels occupy several books, but as he approaches Ithaca, Homer interrupts Book 15 to switch back and forth between him and Telemachus, also headed for home. In cinema, this sort of accelerated crosscutting, often driven by a deadline, has become identified with Griffith’s The Lonely Villa (1909), A Girl and Her Trust (1912), and other Biograph shorts. He lifts the principle of crosscutting to a vast scale in his features. The Birth of a Nation (1915) alternates North and South, home front and battlefront, carpetbaggers and Klansmen in a novelistic fresco.

Crosscutting usually implies some degree of simultaneity. While Telemachus searches for his father, Ulysses leaves Calypso and the suitors run riot in the palace. The notion of actions taking place at more or less the same moment is especially important in chases and last-minute rescues.

As a plot reaches its climax, there can be a sort of site-specific crosscutting too. Once Ulysses and Telemachus have joined forces to slaughter the suitors, Homer’s narration sometimes switches among areas of the fight, as in the battle scenes of the Iliad. While father and son hold off the suitors in the main hall, two servants capture one suitor in a storeroom. We recognize this technique of adjacent alternation when novels and films gather all the major characters in one spot for the climax and shuttles among them.

Crosscutting remains a basic filmmaking tool for most movies on our screens. Where would the Fast and Furious franchise be without it? But some contemporary filmmakers have made fresh uses of the technique. In Inglorious Basterds, Tarantino adopts the big-segment option, alternating lengthy blocks of action before using faster crosscutting when characters converge at the climax. Christopher Nolan has experimented with various tactics, including crosscutting different phases of the same action (Following) and crosscutting among embedded segments, dreams within dreams (Inception).

So there are still lots of options out there to be explored. Just look at some films shown at our Wisconsin Film Festival. Beware, though, of light and heavy spoilers.

 

Attachment plus anxiety

FRantz 500

Frantz (2016).

At one end of the spectrum: Must you always use crosscutting? Wigilia, a charming short feature by Graham Drysdale, suggests not.

It’s built on two Christmas eves a year apart. In the first, a Polish refugee who cleans house for a brusque businessman is alone for the holiday and in his apartment prepares the traditional holiday meal—not for herself but for her absent family. She’s interrupted by the businessman’s vaguely hippy brother, and the two learn about each other as they share the meal. In the second evening, after the businessman has left the apartment to his brother, she returns and they bond more intensely.

Wigilia 400

Apart from an inserted dream sequence, we stay within the apartment. This concentration derives from the production circumstances; Graham explains that he was given a chance to make the film in short order, and to keep it manageable he came up with the idea of limiting the locale. He shot 53 minutes of footage in five days in the apartment, then did the two final scenes in two days. The narrowly focused drama, with many lines improvised, has no need for the free-roaming tactics we associate with crosscutting.

Crosscutting tends to give us a fairly unrestricted range of knowledge; often we know more than any one character. In The Girl and Her Trust, the telegraph operator holding off the robbers can’t be sure that her boyfriend is rushing to her rescue, and he can’t know how close the robbers are to seizing her. Alternatively, when we’re mostly restricted to one character, we don’t find a great deal of crosscutting.

That’s the case in François Ozon’s Frantz, a remake of Lubitsch’s Broken Lullaby (1932), also shown at WFF. Anna’s fiancé Frantz has been killed in the Great War. Out of sympathy his parents have taken her in and treat her as a daughter. But when she sees Adrien, a melancholy Frenchman, haunting Frantz’s grave, she gets curious. Most of the ensuing film is restricted to what Anna learns,

Adrien visits the family. Flashbacks lead us to think he’s what he hesitantly claims to be: a friend of Frantz from prewar Paris. But he has been pressed to tell the parents what they wished to hear. Adrien actually came to the village to beg their forgiveness for killing Frantz on the battlefield, where they met for the first time.  Although we surely have reservations about the sad, apprehensive young man, we don’t learn the truth until Anna does, at about the midpoint of the film’s running time. By this time she has fallen in love with him.

Frantz 2 400

There is some alternation of viewpoint in the film. A few scenes attach us to Adrien during his stay in the village, chiefly when he confronts bitter locals who still consider France their enemy. Still, these scenes don’t give us much direct information about the true backstory. And after Adrien has left and Anna has sought to keep the parents in the dark about the past, we remain attached to her. No crosscutting shows us Adrien’s return to France and his life there. As a result, we’re able to feel curiosity and suspense when Anna decides to track him down. The revelation of his civilian life raises a set of unexpected conflicts.

Both Wigilia and Frantz show that avoiding crosscutting can be a powerful way to keep our attention fastened on characters, the better to let their words and behaviors, as well as their inner lives, get primary emphasis. Crosscutting yields a panorama, while refraining from it can aid portraiture.

 

Crosscutting as usual

Student 500

The Student (2016).

More toward the center of the spectrum lies ordinary crosscutting, the alternation among scenes that provide a broad perspective on the action. In Arturo Ripstein’s Western Time to Die (1965), the plot alternates between scenes featuring the returning convict Juan Sayago and episodes showing the reactions of different townsfolk—chiefly the sons of the man he killed. We also learn of efforts from women in the town to prevent the sons from taking revenge. This “moving-spotlight” narration isn’t perfectly omniscient, though. The plot gradually fills in information about what led up to Sayago’s crime, while revealing that the sons’ mission would amount to avenging a dishonorable father.

A similar sequence-by-sequence approach is seen in The Student, aka The Disciple, a Russian film by Kirill Serebrennikov. A fanatical teenager has become the scourge of the classroom, barraging teachers and pupils with Bible quotations and denunciations of bikinis. His mixed-up fundamentalism, which leads him at one point to challenge evolution by donning a gorilla outfit, is unpredictable and a pure power trip, Biblical bullying.

His mother can’t manage him, the administrators are reluctant to take stern action, and the school priest sees him as a potential recruit to the clergy. Only one teacher, Elena, challenges him with a mix of humor and sympathy. But to combat his increasingly wild behaviors, which include making himself a full-size cross which he can stretch out on, she too sinks into Scripture. She hopes to quote the Bible back at him and dislodge his dogmatism, but she too becomes obsessed and estranges herself from her boyfriend. Meanwhile, Venya gets his one true disciple, a limping underdog, and his campaign against homosexuality, science, and secularism turns violent.

A good part of the narration locks us in to Venya’s Dostoyevskian ferocity, thanks to a restless use of the “free camera” in lengthy following shots. (The film has only about 150 shots in 113 minutes.) But we do range more widely to get a broader view. The moving spotlight shows the mother’s frantic consultations with school officials, and Elena’s clashes with them, as well as with her boyfriend. Still, the climactic scene, which assembles all but one of the characters in a single meeting, has no need of a broader view. Like Wigilia, The Student draws its final power from drilling down into a confrontation around a table.

 

Gaps and folds in time

Killing 500

Killing Ground (2016).

Griffith rang many changes on his last-minute-rescue template, and one of the most startling occurs in Death’s Marathon (1913). A dissolute husband, bored with his life, decides to commit suicide and notifies his wife by phone. She calls the family friend, who races to prevent the death. Surprise: He’s too late.

A hundred-plus years later, crosscutting builds and then deflates suspense in the Romanian film Dogs, by Bogdan Mirica. There are two protagonists: Roman, a city fellow who has inherited his grandfather’s idle farm, and Hogas, the police chief. Both face off against sadistic hoodlum Samir and his thugs. A human foot has popped up (literally, in the first shot), and Hogas tries to trace its owner, while Roman decides how to dispose of the farm. Samir explores other possibilities, none very savory.

Dogs 400

At first we’re restricted to Roman, who is frightened by distant lights and gunshots out on the property, and Hogas, who doggedly pursues his investigation. The alternation between them keeps Samir offscreen for some time. When he surfaces in a suspenseful drinking bout with Roman, and when Roman’s girlfriend comes to pay a visit, the threats start building.

The climax starts out as pure Griffith. Mirica crosscuts Roman’s drive back to rescue his girlfriend, Hogas’ pain-ridden walk to the farmhouse, and Samir’s ominous approach to the isolated woman. Interestingly, the pace of the cutting doesn’t much accelerate in these last moments; there isn’t a lot of alternation, and the emphasis is on prolonged actions (Hogas’ trudging pace, interrupted by coughing up blood, and Samir’s laconic dialogue with the girlfriend).

By the time Hogas arrives, he finds he’s too late. The conventional mystery and suspense of the first stretch are undercut by showing us only the eerie aftermath of a violent climax. Art-cinema norms can de-dramatize crosscutting, but the maneuver remain a revision of what Griffith tried for in 1913.

Three years after Death’s Marathon, Griffith showed the possibility of crosscutting radically different time frames. Intolerance (1916) interweaves four historical epochs while using crosscutting within each one as well. Since then, crosscutting has sometimes been used to juxtapose past and present (The Godfather Part II, The Hours), or alternative futures (Sliding Doors), or a real story and a fictional one (Full Frontal). Interestingly, Griffith is a bit more daring than these directors. These films usually alternate sequences or entire blocks. At first Intolerance does that too, using titles to mark the shift among its four eras. But as the film reaches its climax, Griffith cuts freely from one period to another. These shot-to-shot time shifts, jumping centuries in the burst of a cut, remain an audacious formal discovery.

In all these examples, we’re cued to realize when we move to another period. But Killing Ground, a grueling Australian thriller by Damien Power, doesn’t announce its time-shifting. It exploits our default assumption that crosscutting implies more or less simultaneous action.

At first Killing Ground does give us rough simultaneity, alternating between the yuppie couple, Ian and his fiancée-to-be-Sam, and a pair of gun-loving locals. But then the couple make camp near another family’s tent.

Killing 400

Through careful use of eyeline matches and other continuity cues, the narration welds together actions that are actually taking place at different times. The family’s evening meal and their foray into the woods happen well before Ian and Sam arrive, but the cutting implies that the two groups are living side by side.

Like Griffith, at moments Power shifts between the two periods on a shot-by-shot basis. Small disparities, like a baby bonnet and the placement of the campers’ vehicles, accumulate. By the time Ian follows one of the psychopaths into the woods, we realize that earlier events have been salted through the present-time action, the better to delay revealing the family’s fate. From then on, orthodox crosscutting takes over as Ian runs for help and Sam tries to hold the rampaging peckerwoods at bay.

 

The kids aren’t alright

At the distant end of the spectrum, how about building a whole movie out of full-blown crosscutting? A sustained example at WFF was Bertrand Bonello’s Nocturama. (Major spoilers ahead.)

For the first fifty minutes or so, we follow nine young people silently threading their way through Paris. They ride the Métro, pace along the street, pair up,  separate, crisscross, and assemble at four sites—a line of parked cars, an office building, a Ministry, a statue of Joan of Arc. They’re setting bombs.

We can identify them only through their looks and behaviors. David and Sarah touch fingers fleetingly on a train. We learn from flashbacks that Samir and Sabrina are sister and brother, and their friend is the younger African Mika, all presumably from emigrant families. Flashbacks also show them meeting to plan their action and, once set on course, dance the night away.

There’s an unexpected shooting, but the bombs go off more or less as planned. The group assembles in an upscale department store to meet another confederate, the security guard Omar who will host them overnight. This brief “nodal” moment of unification melts away. Crosscutting follows them as they wander from floor to floor in a parallel to their passages through Paris.

The first section merges the art cinema’s best friend, the prolonged walk, with a thriller-based suspense: we don’t really know what they’re up to until we see a pistol at around 18 minutes and bomb materials somewhat later. The threads knot when we see a quick montage of the bombs.

Bomb 1 400     Bomb 2 400     Bomb 3 400

This fine-grained crosscutting looks ahead to the fragmentary handling of the action in the department store, where the moving spotlight shifts rapidly as the conspirators disperse, assemble in pairs or trios, and disperse again.

Crosscutting is the principal way filmmakers imply simultaneous action, but a lesser option, often favored by Brian De Palma, is the split screen. Bonello uses this device to show the result of the bombings. The shot looks forward to the quiet surveillance-camera display in the security office as the police prowl the shopping aisles. We see the kids moving from quadrant to quadrant, with an occasional flare marking a nearly soundless kill.

Split 1 400     Display 2 400

The terrorists’ motives are barely sketched, and they’re a cross-section of middle-class and working-class kids. Some are unemployed, others have low-end jobs, while others are on track for professional careers. A flashback shows several, perhaps meeting for the first time, while waiting for job interviews. The film’s second large part paints them as victims of consumer lust as they try on upscale fashions and make-up, but the point isn’t hammered home. To some extent they’re just killing time in what they think is a safe house.

Nocturama‘s crosscut climax balances, in more condensed form, the first section, as the conspirators are discovered by the police. At one point, an innocent who has come upon them by accident gets more emphasis than the gang members. His final moments are replayed through multiple viewpoints, as if the stranger’s fate drives home to them what death looks like up close. Soon enough each one will know exactly.

 

It might seem the height of film nerdery to join up films seen at a festival through their different uses of one technique. But is it any more of a strain than those journalistic accounts of how a batch of festival choices reflects The Way We Live Now? Every Berlin or Cannes or Toronto seems to bring forth think pieces looking for a common thread among radically different films, hoping to find today’s social mood in movies begun perhaps years before? Like most zeitgeist readings, they’re pretty easy to whip up.

But technical choices are more concrete than hints of the mood of the moment. Moreover, if you’re interested in cinema as an art, it can be enlightening to reveal the variety of creative options that are still available. The art may not progress, but our understanding of it can. And it’s heartening to find filmmakers refreshing traditional techniques to give us powerful experiences.

Just as important, studying how our contemporaries find new possibilities in something as old as crosscutting can encourage ambitious filmmakers today. The menu is open-ended. There’s always something new, and rewarding, to be done.


We had a wonderful time at this year’s Wisconsin Film Festival. Thanks to all the people and institutions involved, and especially the programmers Jim Healy, Mike King, and Ben Reiser. Each year it just gets better.

Wigilia is currently streaming on Amazon. Nocturama has just gotten a US distributor, the enterprising Grasshopper Film.

Good discussions of interlaced plotting in medieval tales are William W. Ryding, Structure in Medieval Narrative (Mouton, 1971) and Carol J. Clover, The Medieval Saga (Cornell University Press, 1982). Yes, it’s Carol “Final Girl” Clover.

For Tarantino’s use of block construction and time-bending, go here. We discuss Nolan’s penchant for crosscutting in this entry and that one, and at greater length in our e-book on his work.

Safety film 600

Il Cinema Ritrovato: The advantages of leaving home

La fin du jour poster

Kristin here-

David and I are in Bologna for Il Cinema Ritrovato. Once again there is an overwhelming choice of films on offer, demanding a patient acceptance of the fact that one cannot possibly see anything close to everything one wishes. Careful planning can only do so much.

If there is anything I have learned from the films in the first half of the festival, it is that one should not leave home. In the earliest surviving Mizoguchi Kenji film, The Song of Home (1925), a talented but impoverished young man accepts the idea that staying in his village is best for both himself and Japan.

Nearly thirty years later, Girls in the Orchard (dir. Yamamoto Kajiro, 1953), the heroine must choose between going to Okinawa with her fiancé or marrying a man who can help her maintain her family’s traditional pear farm. Naturally, she makes the right choice.

The heroine of Ousmane Sembène’s first feature, the pioneering Senegalese film La noire de … (aka Black Girl, 1966) leaves her home country for France and the better life she dreams of, only to find herself virtually imprisoned working as a maid in Antibes.

The lesson is clear, and yet those of us who have ventured from around the world to Bologna are all the better for it.

 

Color blooms in Bologna

Thief of Bagdad colorful turbans

Color films have always featured on the program at Bologna, but this year various processes are on display in more threads than usual. While the past three festivals have offered a lengthy retrospective of early Japanese found films, this year’s it’s early Japanese color films. There are vintage Technicolor prints in one series, restored color from the silent era in several threads, and eye-poppers like Cover Girl among the restorations being shown off by various archives and labs.

The first screening on the opening afternoon of June 27 was The Thief of Bagdad–not the Fairbanks silent but the 1940 British version co-directed by Ludwig Berger, Michael Powell, and Tim Whelan. I must have been one of the few in the vast Arlecchino theatre who had never seen it, even in a faded 16mm print. Some were there to recapture the fond memories of their youth.

As a “vintage” print, it had an odd history. This was not a vintage re-release print, as some of us expected. It stemmed from the 1990s chemical restoration which was subsequently digitally scanned. The images looked like the Technicolor films of my youth (not quite the 1940s, but at least the 1950s). It was a relatively early film using the three-strip Tech process, which had really only reached its ideal form in Hollywood as recently as 1939, with The Wizard of Oz and Gone with the Wind.

This print had the eccentricities of the three-strip process. Some shots had poor registration, with red and green rims around the characters, while others were in perfect alignment. The matte lines for the numerous fantastical effects (flying horse, giant jinn, flying carpet) were very obvious, and the color changed suddenly for every dissolve. The print was probably not a bad indication of what audiences would have seen at the time.

The design certainly took advantage of the color process, with numerous false-perspective sets and costumes carefully arranged to show off the range of bright hues that Technicolor could achieve (above).

As for the film itself, it is extremely charming without being one of the masterpieces of the era. It suffers from having a bland pair of actors as Ahmad and the princess who loves him, and Sabu is perhaps a trifle too irrepressible as the titular thief, Abu. Miles Malleson, the comic character actor who co-scripted the film, steals the show as a Sultan so obsessed with elaborate mechanical toys that he trades his daughter to the villainous Jaffar (Conrad Veidt, acting rings around much of the cast) for the flying horse. It was an epic in its day and perhaps helped give rise to the many Technicolor fantasies of the 1950s.

A different sort of range was shown off in a program of silent films restored by the EYE Filmmuseum of the Netherlands. These included hand coloring, as in a 1915 short documentary preserved under its English title, Dutch Types, primarily consisting of shots of villages and schoolchildren.

A 1913 Italian film La falsa strada (dir. Roberto Danesi) was a tinted print. It starts off with a familiar situation of an opera singer giving up the stage to live a quiet life on her rich husband’s country estate. One might expect a young lover to rescue her from her boredom, but instead her very lively show-business friends from the city visit and cause the husband to be jealous of the singer’s apparent preference for their company over his. Unfortunately the final reel was missing.

Una-notte-a-Calcutta 300

Even more incomplete was Una notte a Calcutta (dir. Mario Caserini, 1918, right). Only a couple of scenes totaling eleven minutes survive, but they show off the talents of diva Lyda Borelli and suggest that the settings and costumes for this otherwise lost film were impressive.

The emphasis on color promises to continue next year, as with the hints dropped concerning further early Japanese color films to come on a second program.

 

The auteur of the year

Ruggles 3

Following a long-established tradition, the festival includes a retrospective of a Hollywood director, Leo McCarey. Having seen quite a few of the films on offer, I haven’t followed this thread faithfully. I fondly remembered Ruggles of Red Gap (1935) from a single 16mm viewing many years ago, though, and decided to watch it. I was glad I did. For a start, it was a mint 35mm print and a joy to watch. Moreover, I had remembered Charles Laughton’s performance as hopelessly mannered and eccentric. This time I caught many of the subtle gestures and glances that he used to convey the thoughts of a character who, at least in the early scenes, speaks little and then only very formally. The supporting cast is ideal for the witty script that condenses the overly long original novel.

McCarey got his start by directing two-reelers with some of the best second-tier slapstick comics of the 1920s, including Charlie Chase, Max Davidson, and Mabel Normand. One program of three showed off each in turn. The Uneasy Three (1925) casts Chase as an aspiring burglar invading a society party with two partners-in-crime sneaking in by impersonating a trio of classical musicians. Don’t Tell Everything (1927) has Max Davidson marrying a wealthy widow, only to have his obnoxious freckled son (Spec O’Donnell, as always) worm his way into the household by disguising himself as a surprisingly convincing maid. Finally, Should Men Walk Home? (1927) teams Creighton Hale and Mabel Normand in another stealing-a-brooch-from-a-society-party plot. Normand gives a late, great performance. (Imdb lists this as her penultimate role.)

 

The World Cinema Project restores another three

Insiang

I always try to see the latest films restored by the World Cinema Project, which aims to save important movies made in countries that do not have the archives or resources to protect them. This year the films were La noire de …,  Sembène’s first feature, Ahmed El Maanouni’s Moroccan film, Alyam Alyam (aka Oh the Days, 1978), and Lino Broca’s Insiang (the Philippines, 1976).

La noire de … deals with the post-colonialist effects of French rule in Senegal, with the heroine Diouana (below) eager to visit the France of her dreams. Once there, she is never allowed to leave the apartment of the French couple who has employed her; they told her she was to care for their children, but she is relegated to household tasks.

Our friend Peter Rist recalled seeing this film with a color sequence, but this was not included in the restoration. The informative panel introduction to the film, led by Cecilia Cenciarelli of Project, revealed that a sequence showing Diouana’s arrival in Marseilles was shot in color. The idea was to show the heroine’s hopeful view of her new country, contrasting with the black and white of the rest of her film as that hope dissipates. Cenciarelli said that there is no clear evidence that Sembène intended this color scene to be part of the final film. If it survives, it would make a valuable supplement to a future home-video release.

Going from Ruggles of Red Gap to Insiang was an experience in contrasts of the sort one often has here. Insiang is the film’s heroine, a laundress living in a Manila slum. The film was shot in a poverty-stricken area and incorporates many candid shots of children playing in mud and puddles. Much of the action involves shiftless young men who drink and gossip as the women around them do most of the work (above). Against this reality-based milieu, Brocka sets an extremely melodramatic story of Insiang and her mother competing for the affections of the same wastrel. One suspects that Brocka was trying to make his grim film palatable to a broader audience, but the film was a financial failure.

Maanouni took a very different approach for Alyam Alyam. There is a minimal plot about a young peasant earning money to travel to France or the Netherlands for work. This character and his mother and grandfather, Alyaam Alyaamwho strenuously object to be his perceived desertion of them, appear at intervals through the film. Most of the scenes, however, are poetic views of village life, evoking both the back-breaking labor of the countryside and the beauty of its traditions.

In introducing the film, Maanouni said that he wanted to question why Morocco cannot provide the opportunity and incentive to keep young people from leaving. By emphasizing a lyrical depiction of the countryside and the impossibility of earning anything but a subsistence wage, he makes vivid the sad waste of the nation’s potential–a problem that has persisted for decades since the film was made.

 

The unending march of restoration

The one theme that persists from festival to festival is the thread of re-discovered and restored films. The screenings and, increasingly, the panels and lectures on archival methods, reminds us of how expensive and difficult this process is and how much work goes on each year.

The main film I have seen so far among the restorations is Julien Duvivier’s little-known 1939 film, La fin du jour. (A restoration of his more famous Le belle equipe, 1936, was also shown this year.) It’s the story of a group of actors living in a chateau supported by private charities and dedicated to taking care of aging thespians. They play out their individual dramas against the backdrop of a threatened bankruptcy of the home and a dispersal of its inhabitants to various government hospitals across the country.

There are three primary stories. Cabrissade (Michel Simon) maintains his claim to dramatic fame despite having been only an understudy, and that to a star who never missed a performance. Now entering the home and disturbing its equilibrium is Saint-Clair (Louis Jouvet), an unrepentant seducer and liar. Marny (the less famous but excellent Victor Francen) is a successful actor depressed over his wife’s death, perhaps by suicide, after she ran away with Saint-Clair.

There are numerous small plotlines played out by skilled character actors of the era. The ensemble is interwoven in an impressive example of the “Cinema of Quality,” here practiced by scriptwriter Charles Spaak in collaboration with Duvivier. The tale briefly becomes maudlin toward the end but overall is a touching and often funny depiction of old age among a group particularly reluctant to face that time of life.

In the second half of Il Cinema Ritrovato, I’m concentrating on a small retrospective of Iranian cinema of the 1960s and 1970s, as well as a long-awaited restoration of Satyajit Ray’s Apu Trilogy.


David’s book Figures Traced in Light discusses how The Song of Home displays Mizoguchi’s early mastery of Hollywood-style staging and cutting, before he went on to try considerably different techniques.

Thanks to Manfred Polak for a correction regarding La noire de … His blog entry on the film is here.

La noire de ...

La noire de… (Ousmane Sembène, 1966).

An evening with Mr. Smith

Dad's Stick 600

Dad’s Stick (John Smith, 2012).

DB here, from London:

Last week we held another annual meeting of the Society for Cognitive Studies of the Moving Image. I think it’s safe to say that a hell of a time was had by all. I hope to file a report on some of what was said in nearly 100 presentations across 3 ½ days. For now I want just to note a conference highlight that deserves special notice.

J Smith 225English filmmaker John Smith has been working since the mid-1970s, originally in 16mm and now in HD video. On Thursday 18 June, several lucky participants got a very substantial sampling of his career’s achievement in an evening retrospective.

In their visual wit and their emphasis on perceptual shifts, Smith’s films owe something to the international Structural Film trend, but they stand out by their wit and their almost classical beauty. I saw affinities with Hollis Frampton, especially in Associations (1975), in which a test of word-associations slips hilariously out of control. But Smith seems to lack Frampton’s interest in poeticizing scientific and philosophical concepts. The work sampled in our session was characterized by sensuous appeals: precise framing, saturated color, image-sound counterpoint, and richly detailed landscapes. The conceptual element entered as teasing jokes or half-formed narratives.

I had seen three or four before. Hackney Marshes–November 4, 1977 wasn’t in the program, but Kristin and I referred to it in Film History: An Introduction, because of its brash editing patterns. More famous, I suppose, and screened on Thursday, was The Black Tower (1985-87, below), with its oppressive rectangle looming over a striking variety of locales. The film keeps cutting among widely different vistas and neighborhoods, each with the tower visible in middle distance. The associations pile up: state surveillance, a prehistoric monolith, a dark temple, or an alien spacecraft. It was a pleasure to see again.

Altogether different is another of Smith’s most famous works, The Girl Chewing Gum (1976). A busy street is rendered strange by the soundtrack, in which a commentary seems to be directing the whole ensemble of passersby and traffic. Quite soon the voice-over turns into something quite odd. I find a whiff of Monty Python in this charming lesson in image/sound interference. (You can watch it on Vimeo here.)

Blight 500

One of the most striking selections was Blight (1994-96). The demolition of a nearby building becomes, through Smith’s maniacally locked-down camera, a study in abstract composition and eerie movement. My stills can’t suggest the way in which apparently fixed timbers and bricks suddenly shift or fall, moved by no visible hand. The music by Jocelyn Pook further “defamiliarizes” imagery of collapsing walls and exposed wiring. And the tattered man above seems to be watching the whole proceedings.

Flag Mt 500

The program concluded with recent work. Flag Mountain (2010, above) elicited much comment. Fixed stop-motion telephoto shots of a mountain in Nicosia become almost hallucinatory in vibrant HD video and rapid day/night alternations. Dad’s Stick (2012, above) is a lovely tribute to Smith’s father, but it’s also a witty game of form. What seems at first to be a voluptuous abstract painting turns out to be something more mundane, but now mysterious in its accidental beauty. As a bonus, you get a good-natured teasing of art critics’ jargon. Part of Dad’s Stick is here.)

John proved a delightful and modest commenter on his work. He points out that the fixity of each image allows him to treat them as discrete items, like letters of an alphabet, so that the cutting has a crisp force. He’s come to like HD video, but yearns to “get out the old Bolex and shoot one last film.” He declines all the standard labels—avant-gardist, experimental filmmaker, maker of “artist’s films.” He thinks that his films can be appreciated by anybody who’s open-minded, and I think we all agreed. Quiet yet powerful, John Smith’s work seems to me to attest to the continuing relevance of artisanal filmmaking.


You can visit John’s website here. A large selection of his work is available on the three-DVD set John Smith, available from Lux here. The discs aren’t region-coded. Every university should have this set, as well as the Cinema 16 collection, British Short Films, which includes work by John and his contemporaries.

Thanks to Tim Smith, Murray Smith, and John Smith (no relation, none at all) for arranging this wonderful evening.

Black tower 600a

The Black Tower (John Smith, 1985-1987).

David Bordwell
top of page

have comments about the state of this website? go here