David Bordwell's website on cinema   click for CV

Home

Blog

Books

On the History of Film Style pdf online

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

Rex Stout: Logomachizing

Lessons with Bazin: Six Paths to a Poetics

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

Let’s play God, imperfectly: BLOOD SIMPLE on the Criterion Channel

Tuesday | March 23, 2021

Blood Simple (1984).

DB here:

Over the last couple of years I’ve been writing a book on common strategies of popular storytelling in film and other media. I go on to trace how those strategies get worked out in detective stories and thrillers. If you follow this blog, you know that these genres are ones I enjoy and like studying.

So I was happy to offer as one installment of our Criterion Channel series, Observations on Film Art, a short analysis of storytelling strategies in Blood Simple. In it I suggest that although the film has the trappings of a neo-noir–the somewhat downmarket characters, the seedy milieu, the chiaroscuro lighting–in its narrative techniques it’s closer to a Hitchcock thriller. That’s because its manipulation of point of view, one of the resources of popular storytelling, is close to the “partial and misleading omniscience” of the thriller genre and Hitchcock’s narration in particular.

Put it another way. We aren’t restricted to what only one character knows, as in a detective story like The Big Sleep (1946). Instead, Blood Simple steers us selectively from one character to another. So we always know more than any one of them does. That creative choice increases suspense–knowing the dangers that lurk ahead–but it also  summons up an ironic detachment from them, as we watch them make their foolish mistakes. In a word, if you know the film: fish. Here’s another: lighter.

This shifting viewpoint doesn’t give us absolute knowledge, though. There is still some information that slips through the cracks. So we can enjoy superior knowledge in long stretches, while still getting some sudden surprises, or even shocks. (Consider the climax with the perforated wall and the knife at the window.) In their first feature, the Coens prove themselves already fully in control of finely-tuned cinematic storytelling. We’re a step ahead of the characters, but the film is a step ahead of us.

The Coens are also aware of our pleasure in their control. The characteristic Coen awareness, a sly recognition of letting the audience share their power over our access to the story world, is everywhere in evidence. I didn’t have time to mention the shot that everybody remembers, the tracking shot down the bar that simply lifts over the drunk sprawled on the bar top. We become aware of how the movie’s unfolding narration is absolutely ruling how we see this world, and the Coens make a gag out of it. Even the camera-god has a sense of humor.

If you have a chance to watch it, Jeff Smith, Kristin, and I hope you enjoy it–and, of course, the film, which is endlessly rewatchable.


As usual, thanks to the team at Criterion: Peter Becker, Kim Hendrickson, Penelope Bartlett, and their superb postproduction boffins. We recorded the commentary under Covid conditions, with the expert guidance of Erik Gunneson, Meg Hamel, and James Runde.

For more on the Coens’ mastery of storytelling technique, check out the analysis of their fine The Ballad of Buster Scruggs.

You can sample other blog entries on mysteries and thrillers in several entries, in particular our discussions of Hitchcock (of course), and the genre as a whole (here and here and here). An early version of one chapter of my book in progress is devoted to the great Rex Stout. Another chapter will revise what I said about Gone Girl. I also discuss Hollywood’s approach to crime and mysteries in the book Reinventing Hollywood: How 1940s Filmmakers Changed Movie Storytelling.

Blood Simple (1984).

 

A fast-paced cinematic impeachment trial

Tuesday | March 9, 2021

DB here:

Donald Trump did not lose the 2020 U. S. presidential election. He was inaugurated on 4 March. Joseph Biden was arrested and executed.

Or, if you like: Trump will be sworn in on 20 March, the anniversary of the founding of the Republican Party (in Ripon, Wisconsin). Or Trump is President already and working with the military to prepare the mass executions of Democrats. Meanwhile, the Pope was arrested, or will be.

These are among the many lies and delusions  promulgated by fantasists, particularly QOnan (not a typo). But these are also narratives.

Narrative in many manifestations, from jokes to comic books, is one of my keen interests. While mostly studying storytelling in film, I noticed some years back how the term and concept of “narrative” became a part of political discourse. I used that buzzword as a path into analyzing the yarn-spinning around presidential campaigns (here and here) and, inevitably, stages in the fascist coup attempted by the fumbling but indefatigable Trump regime (here and here and here and here). The coup effort continues, now at the level of Republican state legislatures.

The 6 January assault on the US Capitol, an atrocity many of us followed in real time, has been recast as different stories. The versions range from treating it as an escalation of Trump’s seething rallies to brazenly caricatural accounts, notably the version promulgated by my senator Ron Johnson (R-Wisconsin), a man of invincible ignorance. The versions I want to look at are those laid out in last month’s impeachment trial (Trump’s second) by the Prosecutors from the House of Representatives and the ex-president’s legal Defense team.

The trial began on 9 February with consideration of whether the Senate could impeach a President no longer holding office. That issue was settled by a vote that impeachment could proceed. The trial proper began on 1o February. The charge was that Trump incited the riot that led to the invasion of the Capitol on 6 January 2021.

On 10 and 11 February, the House Impeachment managers presented their case. On 12 February, the Trump attorneys presented their defense, after which both sides entertained questions from senators. The final day, 13 February, was taken up with debates about whether witnesses and further testimony would be invited. After that was resolved, the proceedings ended with a vote on conviction. Sixty-seven votes would have been enough to convict Trump, but only 57 votes for that outcome were cast. Trump was acquitted.

Trump’s original speech to his followers is transcribed here. Official transcripts of the Senate sessions are available on the Congressional Record site. A video record of the proceedings, with good quality on the video clips, is on the C-Span site. This also includes a rolling transcript.

Both sides used film to make their arguments. The House team promised what the press called “a fast-paced cinematic case,” and the Defense responded with pieces of cinema of their own. What resources of cinema did they use? And how do the uses differ?

 

Narratio as you like it

In Film Art: An Introduction, Kristin and I contrast narrative form with other options. We suggest that a film (or a literary text or a TV show, etc.) could be organized in various ways. Take whales. Using categorical form, you might provide a taxonomy of whales. Using rhetorical form, you could make an argument about whale conservation. With associational form, you might provide a poetic meditation on whales as metaphors for strength and grace. With abstract form, you might film whales in ways that turn them into pure patterns of imagery and sound. And using narrative form, you might tell a story about how a stubborn young girl helped get a stranded whale back into the sea.

Any given film could combine these principles. An argument for conserving whales could include sequences that give us information about genus and species, while other sequences could poeticize the great behemoths in the Whitman manner, or let us appreciate their abstract beauty. Since we all like stories, there will be a strong temptation to “narrativize” when we can. Even a film that tries to make an argument is likely to include some passages involving agents and actions, goals and obstacles.

A rhetorical film doesn’t have to include a narrative component, but it may well do so. Our example in Film Art, Pare Lorenz’s The River, relies on the classic rhetorical  structure of problem/ solution. But inserted in that structure we find stories about how the Dust Bowl came to be a disaster area and how the government has taken steps to improve things.

A court case is a prototype of rhetoric. Aristotle accorded little importance to storytelling in forensic rhetoric, but later thinkers recognized the powerful role that it plays. Theorists came to call narratio the portion of the argument that tells the story, or rather competing stories, of the matter in dispute.

All of which is a long-winded way of saying that we shouldn’t be surprised that the House Prosecutors and the White House Defense team would float some stories. But what stories were told? How did the advocates tell them? Do the presentations push beyond narrative to other types of form? And what implications do these strategies have for how political actors use moving-image media?

To analyze the cinematic strategies at work, it’s useful to know how they further the broader cases being made. So for each side, I’ll try to show how its use of cinema grows out of its general rhetorical/narrative strategies. Beyond that, we are in a strange re-creation of the situation dramatized in Fritz Lang’s staggering Fury (1936), a film for our moment if ever there was one. There people accused of lynching a man must watch a courtroom screening of their actions during the crime.

     

Needless to say, there was much less self-examination shown by Republican senators when confronted with powerful sounds and images showing the fury they unleashed and did nothing to forestall.

 

Witnesses, plenty, for the prosecution

Analyzing a piece of rhetoric demands that the audience be considered. Since rhetoric is centrally about persuasion, how is the argument shaped to suit the audience’s inclinations?

Almost no one expected that Trump would be found guilty, so were the House’s efforts to mount a case a hollow exercise? I think not. The House advocates, it seems to me, were addressing two audiences: the general public and, to put it grandiosely, history. Representative Jaime Raskin and his colleagues treated the matter as a criminal case, offering a detailed reconstruction of motive, method, and opportunity. Assuming that future analyses would be skeptical and rigorous, looking for holes in the Prosecution account, they tried to make it tight, evidence-based, and the most plausible explanation of the events of 6 January.

Their account can be revised in the light of new information (such as recent findings of who funded the rally and how it was coordinated with the White House). Still, it aspired to be a reliable first draft of the most accurate and comprehensive story of the event. The managers injected emotion into their account by emphasizing the casualties of the melée, the narrow escape that legislators made, and the damage to American ideals when the seat of government was ravaged.

The Prosecution accordingly constructed a complex, time-looping, “novelistic” narrative. Taking advantage of the primacy effect, they laid out their story in a dense fifteen-minute video at the start of the first day, during the argument about constitutionality. I’ll consider that montage shortly, but the crucial point is that it concentrated on the attack, with immersive and brutal footage.

Over the next two days, the House managers spiraled out from the attack footage by means of a plot structure that starts with a crisis. This strategy introduces us to the action just before the story’s climax and then fills in what led up to it.

The Prosecutors’ first step was to establish the story’s antagonist, the “inciter in chief.” Trump was put front and center in a flurry of quotes pulled from moments before, during, and after the raid, ending with him declaring to his followers: “We love you, you’re very special.” Who opposed this central figure? Interestingly, the Protagonists initially cast the Capitol police, particularly the black officers who were stunned by the racism they confronted.

Settling on Trump as the gang leader who fled the scene, the House team set out to prove, through detailed timelines, that the attack was “well-orchestrated.” The presentations were divided neatly into parts and chapters. The sections are mostly chronological, showing events leading up to the crisis, though further flashbacks are embedded. The first full day of the case provided a “moving spotlight” narration on the events, roaming back through time and skipping around in space. Thanks to cellphones, we can get a godlike view of the whole event–a crucial choice of viewpoint for the prosecution. I reconstruct the plot as follows.

The immediate provocation: Three messages (The Big Lie, Stop the Steal, Fight Like Hell) were hammered on after the election and then restated as talking points during the rally.
    Flashbacks [establishing motive and method]:
     Pre-election: Trump announced his resistance to any reports of losing.
     Election: On the big night, he tried to stop the counting and declared that he won.
     Post-election: He refused to concede, fought in the courts, and tried to round up phantom ballots. After failing to win over reluctant senators and he Department of Justice, he saw Pence as his last chance to overturn the electoral results.
When Pence demurred, Trump sought another avenue for action. He invited his followers to come to DC on 6 January to “stop the steal.” “Will be wild!” “He knew what he was doing.”
     The means: Violence. Trump has a history of support of militias and white supremacists (e.g., Proud Boys). His alt-right supporters reciprocated through social media, championing Trump’s Big Lie and declaring 6 January as Armageddon.
[The opportunity:] Back to the 6 January rally, with a split viewpoint: Trump’s speech is intercut with crowd response, using footage from participants’ phones.
Just after the rally: A timeline from 12:20, when the march to the Capitol began, to 2:10, when marchers overwhelmed the police and broke in.
The attack: The timeline continues to the end of the day. Multiple viewpoints provide phone imagery, surveillance footage, and tracking devices to follow the mob’s movements through the building. The assault is crosscut with the responses of police, legislators, and staff. Pence and others were rushed to safety. The mob filled the Crypt, burst into the House, and ransacked the Senate chamber. Outside, rioters boasted of the destruction. “I bled on it.”
Parallel action: “What was happening at the other end of Pennsylvania Avenue?” A new timeline traces Trump’s action from 12:20 at the rally (“I’ll be there with you”)  onward. At 2:38 Trump acknowledged the attack in a tweet. Meanwhile, his staff and members of Congress pleaded with him to call the marauders back. At 6:01 he urged the mob to go home and “Remember this day forever!”
Dispersal of the mob: At 7:00 PM the Senate certified the Electoral College vote.

Thus ended the first day of the House managers’ incitement case. The first stretch of the next day was devoted to replaying events from different points of view, as a novel or film might take us into individual characters’ experience.

The rioters’ testimony: Trump had a bond with the rioters. They asserted that he summoned them. “Our President wants us here.” “He’ll be happy.” His control over them was demonstrated when he asked them to go and they dispersed.
    Flashbacks establishing the power Trump has over his followers:
     Over the years: Trump roused his crowds with insults, making news on TV and nodding to white supremacists. He praised the militias that invaded the Michigan capitol, a “state-level dress rehearsal” for the 6 January siege. Trump did not upbraid the men who plotted to kidnap the Michigan governor; he joked about it at a rally. “This is his essential M.O.”
Trump’s response after the riot: Even after the rampage Trump declared his speech “totally appropriate.” He showed no remorse and failed to condemn the attacks. Even his order to go home was couched sympathetically (“We love you”). His video of 13 January regretting the violence refused to admit that Biden won the election, thus maintaining the Big Lie. Even some Republicans, including former members of his administration, have castigated his failure of duty .
Consequences for the radical right: Most immediately, there was rejoicing on social media (this was “the first stab”) and the riot was proof-of-concept for further missions. There were reports that his followers planned to attack Biden’s inauguration and are continuing to target minority communities. Long-term, this is Charlottesville’s “Unite the Right” rally writ large. Trump has forged a powerful coalition of the fringe groups on the right
Implications for the stability of government: Trump was prepared to sacrifice members of the legislative branch. Through television reports, he knew that everyone, especially Pence and Pelosi, were in mortal danger. While lawmakers were saying goodbye to their families, Trump abandoned his duty. In addition, Trump had no concern for the police or the other staff caught up in the turmoil. Some people were killed and many seriously injured. Ex-service staff called the siege worse than what they had seen in the Middle East.
Implications for national security: Other countries rejoiced at the failure to protect the center of government. (Russia: “Democracy is over!”) Allies were shocked and horrified.
Anticipations of defense arguments (confutatio in the rhetorical trade): This isn’t a Free Speech issue; indeed, Trump launched an attack on the First Amendment. Nor has Trump been denied Due Process.
Conclusion (peroratio): Summing up, with emotional weight added. If this isn’t an impeachable offense, what is?

My outline can’t do justice to the fine-grained detail of the evidence provided in the speeches and the fusillade of images and sounds. I want just to indicate that the argument that a “high crime” was committed and needs to be punished is framed almost completely on the basis of a chronological chain of cause and effect–i.e., a narrative. That narrative indicates motive, method, and opportunity, but these categories of culpability lie beneath the plot we’re given. A man incited a riot by priming the Big Lie, building a violent following, scheduling a provocative gathering during a crucial moment in the transfer of state power, and ranting at a rally.

Beyond the commission of the crime, there is a plea not usually heard in a criminal case. Letting this prominent man off not only sends a harmful message but threatens the future of American democracy.

On the first day, the Prosecutors screened a video that condensed the narrative they propose. While the clip includes some footage not previously shown in the media, it doesn’t include all the new material, some of it quite powerful. Those shots are sprinkled through presentations shown in the following days. The later presentations also recycle many of the shots and sounds presented in this montage, usually to recontextualize them. This montage introduces some of the stylistic methods that would reappear in the days to come.

From Trump’s speech at the rally to his final benediction (“Great patriots. . . .Remember this day forever!”), the video traces a completed arc of action There is no voice-over narrator to guide us, although somber intertitles, white on black, frame the sections. Similarly, there is no music. We hear only the voices and noises in the scenes, although subtitles are brought in to clarify muddy lines and to stress major turns in the action (“They’re leaving!”).

The overall style is that of direct cinema, where the technical roughness conveys the authenticity of events caught on the fly. While some news reportage is included, there’s no editorial comment or punditry. The fact that most of the footage was recorded by the rioters themselves adds conviction. Later presentations will draw on surveillance video to provide a more detached perspective, but the immediate purpose is to immerse us in the participants’ viewpoints. Their blasé attitude toward chronicling their crimes will support the Prosecution’s claim that they believed they had the president’s permission to invade the building and fight for him.

From the start, the video stresses the almost mystical bond between Trump and his supporters. Like Lang’s courtroom scene in Fury, the rally is cut to show both the event and the crowd’s reaction. In eyeline-match editing Trump’s acolytes respond eagerly to his words. After he says go to the Capitol, they reply, “Take the Capitol!” Later his speech is played over shots of their march–the sort of “sonic crosscutting” pioneered by Lang, again, in M (1931). When POTUS says he hopes Pence has the courage to reverse the electoral count, we cut to a shot of Pence. Soon enough the rioters will take up the call: “Traitor Pence!” It’s all part of a single chain of causality.

The Prosecutors anticipated that the Defense would play up Trump’s pro forma aside that the demonstration should be peaceful. But what exactly were the demonstrators supposed to do at the Capitol? To show that violence was really their only plausible option, “Stop the Steal” becomes a unifying thread of the piece. Announced by Trump, picked up by the crowd, and flashed before us in signs and hollers during the insurrection, the phrase suggests not a strategic plan for later primary campaigns but a demand for an immediate tactic, something short and sharp to halt the electoral process.

Crosscutting to show simultaneous events is central to the video’s style. Crosscutting links Trump’s speech to his crowds on the march. It broadens out to trace, for the sake of suspense, many fronts in the assault on the Capitol. While Mitch McConnell rebuts the Big Lie, crowds burst over the barriers and pound on police. Images of the panicked legislators and staff alternate with events elsewhere in the facility, with the danger coming ever closer. Moreover, thanks to some slight sound bridges, the editing gives the impression that the mob is just out of frame in certain shots of the House and Senate chambers.We aren’t far from Griffith’s celebrated crosscutting in The Lonely Villa (1909) and other last-minute-rescue plots.

Just as important, the crosscutting anticipates the parallel timelines that the Prosecution will flesh out in the days to come. Another method, that of split-screen commentary and comparison, is established; it will become useful for yielding the God’s-eye view central to the Prosecution’s case.

     

The video saves the most harrowing footage for its climax, as we see the rampaging mob battling police in the corridors. Ashlii Bobbitt is shot dead. Handheld imagery from the packed crowd records brutality at close quarters. Shots of the raiders hurling themselves forward in unison are intercut with shots of a guard crying out when his head is pinned in a doorway.

     

The epilogue takes us back outside the building. Trump’s video calling for peace is undercut by savage assaults on police on the front steps. Throughout, occasional shots of seething bystanders on the sidewalk have functioned as an ominous chorus (“30,000 guns”/ “Next trip”). Now the video ends on the sidewalk, where an alt-rightist calls for assaults on state governments. A title summing up the seven deaths that resulted from the attack is followed by a shot of Trump’s tweet, bookending the opening, thanking the raiders.

In the days afterward, the Prosecution’s case became an audiovisual one; in the age of Powerpoint, verbal arguments are supported by the evidence of images and sounds. Slides of tweets, Facebook pages, news headlines, and other texts were added to a surge of video clips, along with audio recordings from the beleaguered officers. The direct-cinema quality of the material is amped up by the presentation of body-camera footage from an officer beaten down by the mob.

The fallen-camera convention of pseudo-documentary films (like, say, Cloverfield and Georg) wouldn’t be as powerful if we didn’t know that in violent situations like this, cameras–and camera wielders–do drop. In this instance, the approximate optical POV of the stomped officer makes his attackers seem even more brutal.

The Trump Defense team will claim that the senators know all about what happened and so don’t need to have the entire attack played over and over. But the Prosecutors maintain that in fact we still don’t know a lot, and at the time nobody had a view of the whole drama as it unfolded. That’s the rationale for the timelines, for all the crosscutting and side=by-side juxtaposition of images of things happening at the same time.

This omniscient viewpoint is crucial to the case, and is seen with particular density and vividness during Delegate Stacey Plaskett’s use of visual evidence to complement legal documents. She mobilizes Facebook posts, surveillance footage, audio recordings from barricaded staff, and scale models to trace the rioters’ progress through the building. The section is capped by a confession from perhaps the most famous perp.

These God’s-eye views give a new meaning to the crosscutting initiated in the opening montage, filling in both specific locations and moment-by-moment changes in the rioters’ positions. The bonding of the mob with Trump is recalled when one uses Trump’s nickname, “Crazy Nancy.” And the scrutiny of the footage yields details like the stun gun carried by Barnett, aka Bigo. (Bigo, who “will not back down,” is now apprehensive about his prison stay.) The sweeping range of knowledge is counterbalanced by moments of intimacy, such as the whispers of staffers into their phones.

One advantage of the House managers’ omniscient framing is to mark out two layers of complicity. The parallel timelines make Trump’s actions during and after the assault integral to the act of incitement in the rally. As with an arsonist who sets a blaze and then sticks around to watch and impede efforts to put it out, Trump’s dereliction of duty is presented as of a piece with what Mitch McConnell would later denounce as his “provocation.”

Apart from the feelingful appeals of the succession of speakers laying out the case, the aggressiveness of the action captured in imagery and sound (the thudding doors, the babble of obscenities) and the panicked voices of police and staffers calling for help intensify the emotional dimension of the case.

 

Playing the Trump card

The Defense team were also playing to the general public, particularly the right-wing media. But their chief audience was the Republican senators, who needed a peg to hang their inclinations on. Declining to offer point-by-point rebuttals of the Prosecutors’ case presented a wide-ranging batch of arguments, a buffet of options rather than a tightly focused throughline. There are some sketchy counter-narratives in there, but they’re accompanied by other points, some legalistic, some evidentiary, and some simply suggestive.

The governing emotion, I think, was barely controlled outrage. The House charge, the Defense maintained, reflected a deep hatred of the President and the ordinary Americans he represented. The impeachment was an act of “political vengeance.” While the Defense often said that it would be better if everyone simply calmed down, there was little doubt that the best response to Democratic intemperance was equally fierce counterattack.

Trump’s Defense team relied heavily on arguments of principle. They didn’t try to refute the Prosecutors’ narrative. Indeed, they simply chopped off large stretches of it. They agreed that the sacking of the Capitol was “horrific,” but all the violence there was irrelevant to Trump’s case. The real perpetrators must be rounded up and brought to justice. If he didn’t incite the attack, then he’s not responsible, and all the shocking footage, all the timelines and tracking of the mob, is so much sensationalism.

How to show he did not incite the attack? Mostly by appeal to categorical form. For one thing, the Defense wanted to explore the relevance of two legal concepts. These the Prosecution correctly anticipated in its pre-rebuttal.

First, the Defense team asserted that Trump was denied Due Process on many grounds. Evidence that would be challenged in court was put forth as reliable. Video clips were edited to omit statements in Trump’s favor, and some of the slides of tweets had errors in dating or attribution.

The House evidence of Trump’s encouragement of violence was likewise claimed to be selective and taken out of context. Trump is a man of peace and he has never stoked anger–unlike the Democrats who in many speeches have indulged in hateful rhetoric. Nor is questioning an election result inherently inciting; many Democrats claimed that the 2016 election was stolen.

All this mitigating evidence would have come to light had Trump been given Due Process. Witnesses would have been called; depositions would have been taken; the Defense could have interrogated those accounts. As presented, the case for incitement would not stand up in any court. It is a prime example of the category “Politically motivated injustice.”

Second, Trump was said to be simply exercising his First Amendment rights. Contrary to the opinions of the 140 law professors who declared that Trump’s remarks fail as Free Speech, his statements at the rally are, perhaps lamentably, within the range of current public discourse. His urging his followers to fight simply echoes what Democrats have told their crowds on many occasions.

The Defense videos, fewer in number than the Prosecutors but recycling many shots several times, were designed to drive home these points. Significantly, the first video was designed to establish the “political vengeance” motive.

Apart from ascribing a vendetta to the impeachment initiative, this montage sets up the impromptu category of “Currently acceptable political discourse.” Specifically, the clip plants the “fight” motif that will dominate the later proceedings. It also prepares us for the throbbing music that will underscore the big sequence to come.

The Defense began by bracketing off the entire insurrection, explaining that they wouldn’t be showing footage of it.

They [the Prosecutors] don’t need to show you movies to show you that the riot happened here. We will stipulate that it happened, and you know all about it. 

The Prosecutors’ embrace of the omniscient perspective sought to show that none of the legislators could have “known all about it” then or even now. But the Defense’s dismissal of the entirety of the sacking of the Capitol cleared the way for the case predicated on categorical form. In backing up the argument against conviction, the Defense’s major video sequence assembled samples of prevailing political rhetoric drawn from Democrats’ speeches and statements.


Unlike the Prosecutors, the Defense team relied largely on material from the news media. (Presumably these excerpts don’t count as Fake News.) And, again, the clip is accompanied by throbbing, ominous music evocative of an urban action thriller. If the House case resembled a direct-cinema documentary, the Defense case is stylistically closer to a political ad. All it lacks is a title telling the viewer the sponsor who approved it, and where to send money.

Some of these Democrats’ remarks, taken singly, are pretty abrasive. But although the extremism on display is regrettable, the Defense argued that the Democrats have a perfect right to make them. So does Trump. The Democrats were not telling their crowds to resort to violence, and neither was he.

There is a certain irony in the Trump team’s use of editing to reiterate a concept. It’s a technique we can trace back to other Griffith films, like A Corner in Wheat (1909), but it’s most associated with Soviet filmmakers like Eisenstein and Vertov. The pounding rhythm of “fight, fight, fight” has some of the same force we get in those percussive intertitles in Russian silent classics.

And maybe, the Defense asked, the Democrats aren’t so innocent? A sketchy narrative is invoked in the first sequence above, which arranges statements chronologically to suggest a crescendo of cries for impeachment from the beginning of Trump’s term. The editing of the last montage moves even more strongly toward narrative patterning. Didn’t Democrats’ incendiary rhetoric play some causal role in the summer 2020 civic unrest?

Unlike the Prosecutors, the Defense sequence gives no specifics, no timelines or maps, no link between these sound bites and particular actions in particular places. Many of the Democratic speakers are declaring that they will fight; they aren’t all urging their followers to do so. Unlike the MAGA-blazoned rioters in the Capitol, the street fighters in the videos go unidentified. The episodes of street fighting are a cinematic version of Republicans’ equivalency between Black Lives Matter protests and the assault on the Capitol. This loosely plotted story gives many senators a peg to hang their preferences on.

What happened when the Defense got down to linking up events more exactly? In contrast to the panoramic sweep of the Democrats’ narrative, the Defense scope is very narrow: essentially, just the rally. Apart from some parsing of what Trump meant by “very fine people on both sides” in Charlottesville, there is no mention of his long record of inflammatory remarks. Instead, the Defense constructed a story of what Trump was really asking his crowd to do in that particular speech. They did not try to explicate the many other statements on 6 January, such as the tweet at the top of this section, that congratulate the rioters. The Defense team rewrote the rally as an isolated event, a stand-alone scene whose principal lines demand careful interpretation.

Split screen is used to show the edit points of House videos and to suggest that what Trump said in toto was mitigating.

According to this story, on the day in question Trump was not demanding violence. His call for courage and fighting was simply a request for his supporters to challenge errant Republicans in future elections. If any Trump supporters did engage in the riot, it was because they misunderstood him.

Their error was no greater than that of the Prosecutors, who initially believed that Trump was calling for blood. Once the House managers realized their mistake, Defense attorneys reasoned, they found other pretexts for attacking him. They whipped up the myth of the Big Lie about electoral finagling. Why is it a myth? Because in previous certification ceremonies Democrats accused electors of irregularity as well. They didn’t get away with it because they lacked Senate support and were shut down by the presiding official–in one case, none other than then-VP Joe Biden. Nonetheless, questioning electors’ bona fides, the Defense argued, is not uncommon after a presidential election.

Another, rather sketchy Defense story invoked the possibility that the violence visited on the Capitol was conducted by groups independent of Trump’s control. Some rioters came planning to attack the building and used his rally as a pretext for their mission. They “hijacked the event.” Trump could not incite something that was going to take place anyway. Moreover, some of those people were anti-Trump activists wanting to make mischief. Little evidence of these claims is adduced, but again a proper trial would presumably have brought such information to light.

In all, the people to blame are the Trump followers who misunderstood his message, the alt-right renegades who wanted to make trouble, and the undercover provocateurs from the left. Trump’s remarks, protected by Free Speech, remain distinct from the actions of these troublemakers.

A larger narrative, ultimately more powerful for the public audience, involves motives. First there was Trump’s motive. Trump, on the basis of his speeches, was said to be the most pro-police president in history. He would never encourage his followers to put peace officers in danger. But what about the other side?

Just as the House Prosecutors sought to show that Trump’s desire to retain power made him ever more desperate, the Defense claimed that what fueled this entire proceeding was irrational hatred of a duly elected President. Democrats have “lusted” for impeachment since 2017. The attack on the Capitol gave them the chance to unseat and disgrace him–and keep him from running again. Another montage of quotations, echoing the initial sequence, was invoked. In sum, this is a political trial and the defendant is being railroaded.

The Prosecutors created an intricate narrative centering on Trump and his MAGA army, with those on the other side–from police and staff to legislators both Democratic and Republican–cast as victims. A deadly attack was summoned into existence to keep a man in power. The Defense avoided a long-form narrative but pivoted to categorical organization: Trump’s situation became an instance of denied Due Process, and his tirade exemplified Free Speech in the currently overheated partisan dialogue.

But the Defense didn’t wholly avoid telling a story. It focused on the rally as a single incident that has been misconstrued.  And their broader story, probably the overarching one, pits Trump and his patriotic followers against the vengeful Democratic party and its army of summertime hooligans determined to unelect him, and make him forever unelectable.

 

There’s a lot more to be said about the structure and style of these fraught sessions of the Senate. They will be scrutinized for years by students of law, politics, social change–and, I hope, media. We can bring to the table some tools that allow us to analyze how traditions and conventions of cinematic forms and styles interact with the exigencies of occasion and ideology. They alert us to the ways in which our buttons are pushed and our thoughts and feelings are steered this way and that.

I should say, in case you’re wondering, that I find the Defense case preposterous–sophistic, evasive, streaked with irrevelancies and contradictions. That’s because I side with Aristotle’s view that rhetoric should be a packaging system for sound thinking. Logic and science are truth-tracking, whereas rhetoric makes no discoveries. It is a skill set, a tool kit for presenting ideas and opinions in ways that people will accept. Rhetoric is needed because a lay audience needs help in grasping the premises and conclusions of specialized inquiry. That means, for Aristotle, rooting arguments in not only plausible inference and ethos (trust in the speaker’s moral character, as displayed in the speech) but also appropriate emotional appeals that will engage the audience in the pursuit of understanding.

Ideally, the ideas and opinions set forth are the ones with the best rational and empirical backing. But they may not be. Prejudice, stereotypes, and quick-and-dirty inferences are the rhetorician’s stock-in-trade. Still, we can be on guard against weak rhetoric. Educators could go back to teaching high schoolers the fallacies of informal reasoning; a huge swath of political babble, including the Trump Defense case,  exemplifies tu quoque and ad hominem. And when films and videos take rhetorical form, our skepticism can be reinforced by awareness of how movies work. Studying film can alert us to the ways we can be gulled, blinded, or reinforced in our own prejudices.

It can also show how a rhetorical case can rest on a foundation of more or less plausible explanation. That, I think, is where the Prosecutors’ case leads us, thanks to its strict structure of likely cause and effect. But no less than the Defense, the House managers used the resources of the medium to quicken the impact of the documents they assembled. Fortunately, thanks to modern media, those documents are themselves vivid records in the moment. They will remain, after many viewings for decades to come, powerfully convincing–at least, to people of open mind. The question is how many of those people will still be around.


Geoffrey O’Brien has an engaging discussion of the ways current conspiracy theories echo those of earlier eras in “Hitler in Antarctica” at the New York Review of Books site.

I discuss the role of the crisis structure in classical American film in Reinventing Hollywood: How 1940s Filmmakers Changed Movie Storytelling.

It may be relevant that I thought and still think that Bill Clinton, a man utterly lacking in personal honor, should have been convicted in his 1999 impeachment exercise.

P.S. 21 March 2020: The New York Times has assembled an astonishing montage of mostly new footage of the combat outside the Capitol, showing the police trying to hold the insurrectionists back. Twelve Republican representatives voted against awarding Congressional medals to these officers. One of those opposing the legislation, Rep. Bob Good, was discovered to have a connection to the riot when it was revealed that his district’s party director and her husband were part of the crowd converging on the facility.

     

Left: Sen. Josh Hawley (R-MO) in the Senate Gallery during the impeachment trial; drawing by Art Lien.  Right: Extract from Prosecutors’ video montage. The speaker says: “There’s gotta be something in here we can fucking use against these scumbags.”

Historical film colors: A guest entry from Barbara Flueckiger

Sunday | February 21, 2021

Trois couleurs: Bleu (France/Poland/Switzerland 1993, Krzysztof Kieślowski). Credit: Library of Congress. Photograph of the Agfa Gevaert safety print by Barbara Flueckiger.


Kristin here:

To the general film-going public, old films are in black-and-white. They may be vaguely aware that before The Wizard of Oz and Gone with the Wind, color film was invented.

The history of film color, however, is vastly more complicated than that. Prof. Barbara Flueckiger, of the University of Zurich, has devoted much of her career to studying that history. With Eva Hielscher and Nadine Wieylisbach, she co-edited the 2020 collection, Color Mania: The Material of Color in Photography and Film (Zurich: Eds. Lars Müller/ Fotomuseum Winterthur). Barbara has also spent the past decade leading a team who have created a recently inaugurated and invaluable website that acts as a boundless resource for information on color processes.

We are delighted that Barbara has accepted our invitation to write a guest blog entry for us. She describes the website and gives a succinct outline of the history of film color, loaded with beautiful illustrative frames. Most of these were taken from original archival prints that reveal how seldom–especially in this age of digital home video–we see color films as they looked when they were released.

 

Barbara Flueckiger

From their earliest days, films were colored. During the first three decades, most color imagery was obtained by applying dyes to black and white prints, either by hand, through stencils, or as tinting and toning of the filmstrips. From the beginning, however, many ideas emerged to capture colors directly on film as so-called mimetic colors. That could be done either by optical and mechanical means, such as colored rotary filters, or by chemical interventions, often in combination with optical configurations of cameras. Several hundred analog color processes and film stocks were invented in the first 100 years of film history. Many of them were never successful commercially.

Ali Baba et les quarante voleurs (France 1902, Ferdinand Zecca). Credit: BFI National Archive. Photograph of the stencil-colored nitrate print by Olivia Kristina Stutz.

This history is largely unknown to the general audience as well as to many film scholars and historians.

To close this gap in our knowledge, in 2011 I started to develop the Timeline of Historical Film Colors, a comprehensive web resource. I wanted to document the development of film colors from their prehistory in still photography in the 19th century to the latest developments in the analog domain. As of 2021, the platform contains hundreds of primary and secondary sources, patents, links, selected analyses, physical measurements and downloads, as well as more than 23,000 photographs of historical film prints and negatives. These items provide film historians, researchers, archivists, curators, film restorers and students easy access to a vast array of information. A tagging system connects the entries, galleries, photos and quotes to an underlying thesaurus containing certain topics, persons, aesthetic concepts, technologies, archives, genres, persons or companies. A comparison function allows side-by-side inspection of different prints of the same film.

 

Why film color?

The comparison function allows the side-by-side inspection of different prints of the same film.

High-resolution photographs displayed in galleries are a central part of the Timeline of Historical Film Colors. Early on I developed a method to photographically capture and document historical film materials in a standardized way. It uses a modular and calibrated camera set-up based on a DSLR camera with a macro lens and remote control from the computer to adjust all the parameters. It is crucial to show the full range of color processes in an aesthetically pleasing way, one that aims at recreating the visual impression on the bench, including the edge information and color distribution in the perforation area. These elements are vital for the identification of film stocks and the genealogy of prints.

These photos allow researchers and students to examine individual historical prints, since they often have to work with less-than-ideal digitizations on DVDs and Blu-rays that are just a faint echo of the historical source material. In recent years this photographic method has been adopted by my teams in the current research projects. Some archives, such as the Academy Film Archive, have started to use the method, and the BFI National Archive and the George Eastman Museum plan to do so soon.

Our modular camera set-up in use at the bench.

During the last years my teams and I visited many archives in Europe, the US and Japan to take these photographs, such as the Harvard Film Archive, EYE Filmmuseum Amsterdam, National Film Archive Prague, Deutsche Kinemathek Berlin, the Academy Film Archive, the Library of Congress, George Eastman Museum, the BFI National Archive, Cinémathèque française Paris, the UCLA Film & Television Archive, Bundesarchiv Filmarchiv Berlin, Museum of Modern Art, DFF Deutsches Filminstitut & Filmmuseum Frankfurt, the National Film Archive of Japan and others.

On the Timeline of Historical Film Colors each contributing archive is represented with a header slide that gives access to the film elements from their collection.

 

The lost colors of film history

Most films produced before the mid-1930s have been passed on in black-and-white prints. It was not until the famous FIAF conference in Brighton 1978 that the colors of the first decades of film history began to attract some attention from insider circles focusing on silent film.

To this day, the lack of awareness of film’s colorful past has persisted. Early applied colors such as tinting, toning, hand-coloring, and stencil-coloring are ephemeral by nature, since each exhibition print was dyed separately, in a variety of shades and hues. Moreover, these prints were produced with highly flammable nitrate cellulose as a base. Many deadly cinema fires in the early decades of the 20th century demonstrated the dangers of nitrate stock. Therefore, many original colored film prints have been hidden in cans sitting on the shelves in archives’ nitrate vaults. These facilities are fitted with special safety measures such as break-off walls and earth dams.

Eventually in the 1950s safety celluloid film stocks replaced nitrate. From that point on, new prints of colored early films were made on safety stock from the black-and-white camera negatives, intermediate negatives, or positive distribution prints. When colored distribution prints were used, the new copies were usually made only in black-and-white.

In the early 1980s a second threat to the history of colors in film became apparent. Martin Scorsese was among the prominent filmmakers and scholars who rang the alarm bell over the fading of so-called chromogenic stocks produced from the late 1930s to the 1980s. Due to the physical decay of mainly the cyan dye in these film stocks, original prints become nearly monochromatic, retaining mainly colors in the magenta to red spectrum. To this day, dye fading has remained one of the most pervasive problems for the search of authentic film colors.

Color fading. Blade Runner (USA 1982, Ridley Scott). Credit: Library of Congress. Photograph of the Eastman Color Print Film by Joëlle Kost.

 

Applied colors

During the first three decades, so-called applied colors dominated. Historians estimate that about 80% of film prints were colored by tinting, toning, or hand- and stencil-coloring.

Tinting means submerging black-and-white film positives into dye baths, so that the prints’ gelatin emulsion acquired a more or less uniform, mostly monochrome color. Tinting can be identified by the inspection of the perforation area that is also uniformly colored. Toning, on the other hand, is a complementary process whereby the silver image is replaced by colored metallic pigments (metallic toning) or dyes (mordant or dye toning). In contrast to tinting, toning leaves the perforation area mostly colorless.

Tinting. Malombra (ITA 1917, Carmine Gallone). Credit: Cineteca di Bologna. Photograph of the tinted and toned nitrate print by Barbara Flueckiger.

Toning. Voyage autour d’une étoile (France 1906, Gaston Velle). Credit: Cineteca di Bologna. Photograph of the toned nitrate print by Barbara Flueckiger.

For these coloring processes the individual prints had to be cut into segments that were then dyed in batches and reassembled into the final distribution print. As a result, individual prints can vary considerably in their color schemes.

Comparison of four differently tinted and toned distribution prints of Das Cabinet des Dr. Caligari (Germany 1919, Robert Wiene). Copyright: Friedrich Wilhelm Murnau Foundation. Photographs by Barbara Flueckiger.

Whether tinting and toning schemes vary due to cultural norms and tastes has remained a topic of debate. To a high degree it is also uncertain who made the decisions about the coloring, except for cases where scripts, production notes, or film negatives indicate the attribution of colors. In addition to colored prints there were so-called copyright books that show the color scheme by single frames attached to the pages of the booklets, deposited at the Library of Congress by distributor George Kleine. Subtle shades emerged that often make it difficult to distinguish between the two, because the black silver image gives way to nuanced interactions with the tinting dyes in middle tones.

Copyright book from George Kleine: Gli ultimi giorni di Pompei (Italy 1913, Eleuterio Rodolfi). Credit: Library of Congress. Photograph Barbara Flueckiger.

In some cases, combining tinting and toning allowed for two colors to appear within a single image.

Tinting and toning combined. Sumurun (Germany 1920, Ernst Lubitsch). Credit: Bundesarchiv-Filmarchiv. Photograph of the tinted and toned nitrate print by Olivia Kristina Stutz.

The range and variety are even greater in the case of hand and stencil coloring. These techniques generally required the application of  up to six dyes on each individual frame, either by hand through tiny brushes or by cut-out stencils. These laborious processes were demanding, given the small image area and the huge number of frames, generally 16 to 18 per second of running time. Hand-colored films show an uneven application of dyes with soft transitions between individual colors. For stencil coloring, each dye necessitated a separate, colorless print from which the stencils were cut out by needles or metallic styluses. As a result, shapes appear more or less sharp-edged. It was a mechanized version of hand coloring that allowed the coloring of feature-length films and higher numbers of distribution prints. Over the years, improved techniques were introduced to transfer the shapes from projected magnifications onto the film prints with the help of pantographs.

Hand coloring: Métamorphoses du papillon (France 1904, Gaston Velle). Credit: Library of Congress. Photograph of the hand-colored nitrate print by Barbara Flueckiger.

 

Stencil coloring. Cyrano de Bergerac (Italy/France 1923, Augusto Genina). Credit: EYE Filmmuseum Amsterdam. Photograph of the tinted, toned and stencil colored nitrate film by Barbara Flueckiger.

Needless to say, stencil and hand coloring were reserved for more ambitious or luxurious films. However, they also allowed for the creation of a higher reality effect in documentaries, travelogues, or fashion films by anticipating the development of mimetic colors. Exotic places, ethnicities, or historical settings were among the prevailing topics of stencil-colored films.

Documentary. La mangouste ou rat des pharaons (France 1914). Credit: Cineteca di Bologna. Photograph of the stencil-colored nitrate print by Noemi Daugaard.

 

Fashion film. Modeflitsen (France 1918). Credit: EYE Filmmuseum Amsterdam. Photograph of the stencil-colored nitrate print by Bregt Lameris.

 

Travelogue. Coiffures et types de Hollande (France 1910, Alfred Machin). Credit: Cineteca di Bologna. Photograph of the stencil-colored nitrate film by Barbara Flueckiger.

In fact, the richness and scope of stencil-colored films can be fascinating to the modern viewer. That holds true for both the bolder color in the first decade of the 20th century or the more nuanced pastel shades that became increasingly prevalent in the 1920s.

Bold colors in early film. L’Amour d’esclave (France 1907, Albert Capellani). Credit: Library of Congress. Photograph of the stencil colored nitrate film by Barbara Flueckiger.

 

Subtle pastel shades in the 1930s. Elstree Calling (Great Britain 1930, André Charlot; Jack Hubert; Paul Murray; Alfred Hitchcock ). Credit: BFI National Archive. Photograph of the stencil colored nitrate print by Olivia Kristina Stutz.

A special case of applied colors is the Handschiegl process, a printing process developed by Max Handschiegl and Alvin Wyckoff, often used in Cecil B. DeMille’s films, especially for title cards. It produces highly detailed and precise colors with stunning effects.

Handschiegl. Joan the Woman (USA 1916, Cecil B. DeMille). Credit: George Eastman Museum. Photograph of the tinted, toned and Handschiegl nitrate print by Olivia Kristina Stutz.

 

Mimetic colors

Already in France in the 1860s, Charles Cros and Louis Ducos du Hauron separately wrote descriptions of many of the principles for achieving mimetic colors in still photography. As it turned out, however, it was a much more demanding task to develop solutions for moving pictures. Some of the problems related to the high throughput during projection of 16 or more single frames per second. Other problems resulted from much higher requirements for image size on the big cinema screen, where resolution and registration were paramount. Due to the rapid succession of frames necessary for the illusion of movement, minute deviations occurring between frames created disturbing amounts of flicker or color fringing. Contemporary commentators often labeled the result as “color bombardment” that caused “eye strain”.

To this day, mimetic colors combine two to four color components either in additive or subtractive admixtures. In the 19th century their development followed psychophysical insights into the human visual system by Thomas Young and Hermann von Helmholtz. They showed that color impressions are the result of physiological sensors in the human retina sensitive to three different spectral ranges of the visible light.

 

Additive colors

Additive admixtures operate with colored light where the sum of the three additive primaries red, green and blue results in white light. The earliest attempts to create colors on the screen by optical means employed additive principles by rotary filters in front of the camera and projector respectively. These included the three-color Turner Lee and the most successful additive two-color process Kinemacolor.

Rotary filter in front of the Kinemacolor projector used for David Cleveland’s and Brian Pritchard’s reconstruction. Credit: Brian Pritchard.

 

Kinemacolor positive from the Kodak Film Samples Collection. Credit: National Science and Media Museum Bradford. Photograph by Barbara Flueckiger in collaboration with Noemi Daugaard.

In Kinemacolor, rotary filters in red and green spinning in front of camera and projector recorded and transmitted the color information by temporal synthesis. The impression of color was created in the eyes of the spectators. Based on contemporary reports and digital reconstructions, the poor quality and limited color spectrum were readily apparent. Due to the temporal shift between the two successive frames with the red and green color separations, Kinemacolor and all processes operating by the same principle created color fringes and a headache-inducing amount of flicker.

Mroz Farbenfilm. Urlaubfarbenfilm F. Apfelthaler (AUT 1932, Friedrich Apfelthaler). Credit: Österreichisches Filmmuseum. Video and reconstruction by Giorgio Trumpy, David Pfluger and Martin Weiss.

Attempts with temporal synthesis were followed by additive processes that employed spatial synthesis by the application of beam splitters. In this configuration, up to three color records were taken through filters simultaneously to eliminate temporal parallax. But this approach introduced spatial parallax instead, and this arrangement could also create color fringes by poor registration. Gaumont’s Chronochrome process was certainly the most convincing attempt to combine three color separations with this principle.

Gaumont Chronochrome positive from the Kodak Film Samples Collection. Credit: National Science and Media Museum Bradford. Photograph by Barbara Flueckiger in collaboration with Noemi Daugaard.

A third type operating with additive admixtures of light are the so-called screen processes. Additive colors are mixed, either by random, small-scale mosaic image elements or by organization into lines. Color impressions in these systems result from the fusion of the individual dots or lines into red, green and blue in visual perception. The effect resembles that of pointillist paintings where colors are divided into small dots.

Lenticular screen processes, by contrast, combine tiny lenses imprinted onto a black-and-white film strip with colored filters in front of the camera and projector.

Kodacolor lenticular filter for the projector. Lichtspiel/Kinemathek Bern.

Mosaic screens created by colored potato starch were popular in still photography with the Lumière brothers’ Autochromes, introduced in 1907 and widely used by professionals and advanced amateurs. The principle was later adopted in the Cinécolor process for color film but failed due to the uneven distribution of the starch particles. Among the line screen processes, Dufaycolor was the most successful one, widely used in documentaries and famously in Len Lye’s experimental films with direct animation painted directly onto the film strip and then captured and distributed on Dufaycolor film stock. Apparently, Lye was not convinced by the somewhat muted color palettes of the process.

Stereoscopic Autochrome Lumière. Exhibition Color Mania – Materiality of Color in Photography and Film, Fotomuseum Winterthur, September 7 to November 24, 2019. Photo by Barbara Flueckiger.

 

Photomicrograph of Cinécolor (20x). Credit: Photomicrograph by Silvana Konermann.

 

Cinécolor sample. Credit: Gert Koshofer Collection. Photograph by Barbara Flueckiger.

 

Photomicrograph of Spicer-Dufay, early Dufaycolor (20x). Credit: Photomicrograph by Silvana Konermann.

 

Dufaycolor. A Colour Box (Great Britain 1935, Len Lye). Credit: BFI. Photograph of the Dufaycolor di-acetate print by Barbara Flueckiger.

Lenticular films such as Kodacolor were also mostly used for home movies with the exception of Thomson color for Jacques Tati’s Jour de fête (France, 1949).

None of the additive principles proved to be successful for the long term. Many of them required special installations in the cinema, and most delivered poor results, most notably dim images.

 

Subtractive colors

Finally, subtractive admixtures became the norm. The three primaries cyan, yellow and magenta filtered the light, with black being the sum of these three subtractive colors. Two or three differently colored emulsion layers are attached to the film base, on one side or both sides of the film.

Most early two-color films were using double coated film stock. The earliest one was Kodachrome Two-color developed in 1915, presented in 1916 with the short film Concerning $1000, but mostly in use in the 1920s for fashion films, for the dance film The Flute of Krishna (USA 1926) by choreographer Martha Graham and for the experimental film [Kaleidoscope] by Loyd A. Jones. Kodachrome Two-color film was shot through a beam splitter and combined two emulsions in orange-red and bluish green on either side of the film carrier, with beige, brown and golden tones in the spectrum between the two color components.

Kodachrome Two-color Test Shoot No. III (USA 1922, Anonymous). Credit: George Eastman Museum. Photograph of the Kodachrome two-color double coated stock by Olivia Kristina Stutz.

 

Kodachrome Two-color. [Kaleidoscope] (USA ca. 1927, Loyd A. Jones). Credit: George Eastman Museum. Photograph of the Kodachrome two-color double coated stock by Barbara Flueckiger.

A large range of subtractive two-color processes on double-coated stocks emerged in the 1920s and the first half of the 1930s. These included Multicolor, Polychromide, Sirius Farbenfilm, Sennett Color, and Ufacolor. Many of them were short-lived and tested only with short films or commercials. Even failures, however, can help us understand the technical development and the basic principles applied. Prizma II, one of the many processes invented by William van Doren Kelley, was more widely used for travelogues produced by the Prizma Company, for commercials, and for the first feature film in color, The Glorious Adventure (Great Britain,1922, J. Stuart Blackton). At splices the two color components orange and cyan become visible. Overall, Prizma II has a pale and often dirty-looking appearance with occasional patches in orange-red popping out on flowers, fruit or costumes.

Sirius Farbenfilm. [Farbfilmversuche] (Germany 1920s or 1930, Ludwig Horst; Hans Horst). Credit: Deutsches Filminstitut DFF. Photograph of the Sirius color nitrate print by Barbara Flueckiger.

Multicolor. Credit: Gert Koshofer Collection. Photograph by Barbara Flueckiger.

 

Prizma II. The Glorious Adventure (Great Britain 1922, J. Stuart Blackton). Credit: BFI National Archive. Photograph of the tinted and Prizma II nitrate print by Olivia Kristina Stutz.

In the 1930s subtractive processes turned to three colors, most famously with Technicolor No. IV and the subsequent Technicolor No. V, which was printed from chromogenic camera negatives.  Founded in 1915, the Technicolor Company went through many failures and set-backs, with the exception of a short color rush in the late 1920s with the two-color dye-transfer process Technicolor No. III. Following the series Great Events with 12 short films produced by the Technicolor company to establish the process, mostly musicals and a few other genres exploited the two-color process during this short boom. But some of them are highly remarkable, with sophisticated camerawork by Technicolor’s own cinematographer Ray Rennahan, including the musical Whoopee! (USA 1930, Thornton Freeland) choreographed by Busby Berkeley, King of Jazz (USA 1930, John Murray Anderson, Pál Fejös), Doctor X (USA 1932, Michael Curtiz), and Mystery of the Wax Museum (USA 1933, Michael Curtiz).

Technicolor No. III. Doctor X (USA 1932, Michael Curtiz). Credit: UCLA Film & Television Archive. Photograph of the Technicolor No. III dye-tranfer nitrate print by Barbara Flueckiger.

 

Technicolor No. III. King of Jazz (USA 1930, John Murray Anderson; Pál Fejös). Credit: Library of Congress. Photograph of the Technicolor No. III dye-tranfer nitrate print by Olivia Kristina Stutz.

While the technologies applied in Technicolor’s various processes changed considerably over the years, the beam splitter was one of the few constants. Both in Technicolor No. II and III, a beam splitter separated the two color records and captured them mirrored upside down on one black-and-white negative. The bulky and heavy Technicolor No. IV camera recorded the color separations on three black-and-white 35 mm negatives. From these three negatives matrices were produced as wash-off reliefs, ready for the dye-transfer of the three primaries onto the positive print. The result was a series of color images, along with the frame lines and the soundtrack as silver images.

For almost two decades Technicolor dominated the market for high-quality color films. Part of its success was due to a comprehensive package that included the camera, specialized cinematographers, and all the lab works executed exclusively in Technicolor’s own plants. One of the building blocks of Technicolor’s long-term dominance, however, was the so-called Color Advisory Service, famously led by color consultant Natalie M. Kalmus. She defined aesthetic guidelines for film productions shot with the process, informed by color norms related to concepts of “elevated taste,” located in a broader cultural context with references to the concept of “color consciousness.”

Technicolor No. IV. The Life and Death of Colonel Blimp (Great Britain 1943, Michael Powell; Emeric Pressburger). Credit: BFI. Photograph of the dye-transfer nitrate print by Barbara Flueckiger.

 

Technicolor No. IV. Blood and Sand (USA 1941, Rouben Mamoulian). Credit: BFI. Photograph of the dye-transfer nitrate print by Barbara Flueckiger.

 

Technicolor No. IV. Gentlemen Prefer Blondes (USA 1953, Howard Hawks). Credit: Library of Congress. Photograph of the dye-transfer safety print by Barbara Flueckiger.

Despite all the efforts to control the color schemes, people often associate Technicolor with highly saturated, deep colors. On close inspection in our detailed analyses of color films, however, we have observed that many films adhere to the rules with mostly restrained color schemes and unsaturated backgrounds to guarantee optimal figure/ground separation. But there are also deviations from these self-imposed norms, surprisingly clashing hues even in films produced with Natalie M. Kalmus as color consultant.

Moreover, there is a great variability of different looks and color applications during the almost two decades. Individual color aesthetics were related to personal styles of cinematographers, directors or production companies, genres or changing preferences in fashion and design, and changing color compounds and recipes employed in the process. Technicolor’s idiosyncrasies – what we perceive as typical “Technicolor look” – are mostly due to the dye-transfer process itself. Pasty, dense colors in patchy structures create an almost opaque appearance on the screen, an effect somewhat like oil paint. When we work with the film elements on the bench in archives, we not only have to increase exposure considerably due to the density of the film stock, but we also notice the color layer’s almost sticky viscosity, often visible as a relief on the surface.

Compared to Technicolor, Gasparcolor produced much more saturated, brilliant and luminous colors. In fact, the process, developed in the early 1930s by Hungarian emigré Béla Gaspar in Germany, was possibly the most advanced and complex process at the time. In its principle–the silver dye-bleach process described by Raphaël E. Liesegang in the late 1890s – the silver acted as a catalyst for the local destruction of the dyes embedded in the three emulsion layers on the two sides of a reversal positive. It is thus a chromolytic reversal process. Due to the political circumstances during the Third Reich in Germany, Gaspar eventually had to flee.

Like Technicolor Gasparcolor required the recording of three color separations on black-and-white negatives. Since most of the Gasparcolor films were animations, these separations were captured in succession on adjacent film frames but could of course also have been shot through a beam splitter similar to Technicolor No. IV. In fact, only one documentary is widely known, Colour on the Thames (Great Britain, 1936), shot by Adrian Cornwell-Clyne. Among the films produced with Gasparcolor are famous avant-garde experimental films by Oskar Fischinger, Hans Fischinger, and Len Lye. Gasparcolor prints can easily be identified by the black perforation area and the colored soundtrack.

Gasparcolor. The Ship of the Ether (Netherlands 1934, George Pal). Credit: BFI National Archive. Photograph of the Gasparcolor nitrate print by Barbara Flueckiger.

Gasparcolor. Colour on the Thames (Great Britain 1935, Adrian Klein). Credit: BFI National Film Archive. Photograph of the Gaspar color nitrate print by Barbara Flueckiger.

Gasparcolor. Rainbow Dance (Great Britain 1936, Len Lye). Credit: Museum of Modern Art. Photograph of the Gaspar color nitrate print by Barbara Flueckiger.

Both Technicolor and Gasparcolor prints stored in archives are in remarkably good shape, due to their stable colors. Ironically, chromogenic film stocks, the technical principle that ultimately won the competition and became the new standard, had the least stable dyes. Chromogenic means that the dyes need to be developed after exposure. Embedded in the emulsion of a single strip of film stock are three or more layers. These layers are sensitive to different spectra. All contain silver halides and the color-forming substances, so-called dye couplers that are subsequently developed into dyes. In a second stage the silver is bleached out and leaves the color information in the form of finely dispersed dye clouds in the three or more emulsion layers in cyan, magenta and yellow. The result is a highly translucent, glowing image whose fine-grain structure depends on the speed of the film stock. The slower the speed, the finer the grain.

In contrast to Technicolor the shooting of the chromogenic film could be done with normal cameras on one negative or camera reversal. Chromogenic films increasingly became the norm, starting with Agfa’s first negative-positive process Agfacolor. Emerging in the late 1930s, Agfacolor was promoted by German propaganda in a bid to counteract Technicolor’s dominance. Agfacolor had particularly soft colors in the pastel range with a typical, slightly darkened orange-tomato red. Difficulties in the blue range produced turquoise shades that become quite apparent in skies. Greens had a tendency to look brownish or blackened; shadows had a greenish tinge. Chromogenic multilayer film stocks were incredibly difficult to balance and to produce, requiring a high level of knowledge in physics and chemistry.

Agfacolor. Münchhausen (Germany 1943, Josef von Báky). Credit: Copyright Friedrich Wilhelm Murnau Foundation. Bundesarchiv Filmarchiv. Photograph of the Agfacolor safety print (acetate) by Barbara Flueckiger.

Agfacolor. Opfergang (Germany 1944, Veit Harlan). Credit: Copyright Friedrich Wilhelm Murnau Foundation. Filmmuseum Düsseldorf. Photograph of the Agfacolor Safety Print by Barbara Flueckiger.

Agfacolor. Grosse Freiheit Nr. 7 (Germany 1944, Helmut Käutner). Credit: Copyright Friedrich Wilhelm Murnau Foundation. Bundesarchiv Filmarchiv. Photograph of the Agfacolor nitrate print by Michelle Beutler.

 

Agfacolor. Der schweigende Stern (German Democratic Republic 1960, Kurt Maetzig). Credit: Bundesarchiv Filmarchiv. Photograph of the Agfacolor safety print by Josephine Diecke.

After World War II ended, the Allies were able to exploit German color-film patents. The result was the appearance of Fujicolor, Eastman Color, and many derivatives, such as Ferraniacolor, Ansco Color, and Sovcolor. The worldwide adoption of color in film production soon followed.

Sovcolor. Ivan the Terrible, Part II (Russia 1958, Sergei M. Eisenstein). Credit: Museum of Modern Art. Photograph of the Sovcolor safety print by Barbara Flueckiger. (The film was shot in the 1940s on captured Agfacolor stock, but the delay in the release of the film until 1958 meant that distribution prints were on Sovcolor stock.)

 

Fujicolor. Matador (Spain 1986, Pedro Almodóvar). Credit: Library of Congress. Photograph of the Fujicolor safety print by Barbara Flueckiger.

 

Eastman Color. Aliens (USA/Great Britain 1986, James Cameron). Credit: Academy Film Archive. Photograph of the Eastman Color Print Film Type 5384 by Joëlle Kost.

 

Eastman Color. Gattaca (USA 1997, Andrew Niccol). Credit: Library of Congress. Photograph of the Eastman EXR Color Print Film Type 5386 reference print by Barbara Flueckiger.

A plurality of styles emerged, less defined by technical limitations than by cultural contexts and individual preferences of filmmakers, art directors, costume designers, and cinematographers. Color aesthetics in film are not only created by hues, color schemes, and color contrasts, but also by lighting styles, by material properties of surfaces and textures, by depth of field, image composition, and by movement. The combination of these factors influences the image’s figure-ground relationships.

In the course of our research, we investigated a large corpus of more than 400 films – mainly from 1895 to 1995 – with a computer assisted workflow. A video annotation software has been developed based on our approach since 2017, when we figured out that tools available then were not well suited to the detailed annotation and visual analysis of film (color) aesthetics. The visual analysis and annotation software VIAN has been created by Gaudenz Halter in collaboration with the Visualization and MultiMedia Lab of the University of Zurich. The tools enable researchers to create detailed analyses including figure/ground separation and a large range of visualizations that make diachronic developments immediately evident or support the testing of new hypotheses.

Video analysis and annotation software VIAN, developed by Gaudenz Halter. User interface.

A deepened understanding of color film technologies and aesthetics is an essential prerequisite for the scientifically sound digitization and restoration of color films, which is one of the most pressing topics today and therefore remains at the center of our research activities.


Acknowledgements

I would like to express my immense gratitude to Kristin Thompson and David Bordwell for publishing this blog post and for all the inspiration that guided my research.
A huge thank you to my teams ERC Advanced Grant FilmColors, SNSF Film Colors. Technology, Cultures, Institutions, ERC Proof-of-Concept VeCoScan.

Special gratitude is dedicated to all the film archives with their wonderful collections.

This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme, grant agreement No 670446 FilmColors.

Madalena, Rosalind, and Suzanna: More Rotterdam revelations

Sunday | February 7, 2021

Madalena (2021).

DB here:

A mixure of moods and tones for our final communiqué from the International Film Festival Rotterdam. Its fiftieth year has been a lively one.

 

The Madalena mystery

Madalena (2021).

In earlier entries (especially here) I’ve noted that the thriller genre is well-adapted to festival circulation. It doesn’t require the budget of a blockbuster. It can attract major actors who want tricky parts to play. It can be shot on contemporary locations. And the appeal to suspense and surprise fits comfortably with edgy narrative strategies favored by art cinema. At the limit, a filmmaker can arouse our thriller appetites and then try a bait-and-switch that not only warps the genre’s conventions but sets us thinking.

The Brazilian film Madalena, by Madiano Marcheti, starts as a classic mystery. In a vast field of soy, reas stalk gracefully as a monstrous pesticide-sprayer grinds toward them. But among the rows lies a corpse.

What follows is more fractured and prismatic. A first section attaches us to Luci, a friend of Madalena’s who works as manager of a club. She also picks up work dancing for TV commercials, one set in that very acreage. Then we follow Cristiano, whose father owns the land and demands he hustle to harvest. A third section takes us with trans woman Bianca and her girlfriends, who sort through Madalena’s belongings before setting out for a day of driving, swimming, gossiping, and teasing one another, the memory of Madalena never far from their thoughts.

Marcheti skips some of the standard scenes. We never see the police investigation, or even the discovery of the body.  The crime plot has been a pretext to reveal a cross-section of life in the community, from the wealthy farmers to the cottages where the staff live. The resolution shifts the question of who did it to the broader impact of the death, and how it stands for a horrifying statistic: Brazil has the world’s biggest murder rate of transgendered people.

Throughout, sexualization of bodies is a central motif. Luci and her posse hang out at curbside, Bianca and her posse turn tricks and find boyfriends, and Cristiano, after sizing up the crowd at Luci’s bar, winds up dancing with himself in mirror reflection.

To say much more would spoil things, so I’ll just note that this story is filmed with a pictorial intelligence that one seldom sees these days.  The imagery of the soy fields is at once magnificent and ominous. Drones hover over it like birds of prey, and its horizon haunts the people’s lives.

     

     

Overwhelming as the landscape is, it doesn’t blot out the characters’ routines and the crises that disrupt them. Moving from Luci’s aimless days and nights to Cristiano’s panic to Nadia’s quiet  tribute to Madalena, a locket set adrift in the stream that runs along the field, the film pauses for intimate moments. It reminded me a bit of Varda’s great Vagabond (Sans toit ni loi), in which an enigmatic figure’s fate charts the range of human indifference, but also affords glimpses of sympathy.

An informative discussion of the film with Marcheti is provided by IFFR here.

 

As we too like it

As We Like It (2021).

This movie saw me coming a mile away. It does for As You Like It what Lurhmann did for Romeo and Juliet, but to an Asia-pop beat. Four romantic couples lose and rediscover one another in a magical milieu–not the Forest of Arden (currently under corporate development) but a district of Taipei with no Web connections.  In Heaven, a sign informs us, there is no Internet.

Accordingly, people must deliver messages in person, seek out each other by dint of shoe leather and motorbikes, and actually meet face to face. So Rosalind’s quest to find her father the Duke (a genial tycoon) intertwines with Orlando’s search for her. But of course she’s disguised as a boy and aided by Celia, a fortune-teller who’s the dream girl of Orlando’s sidekick Dope.

The film’s world is maximum kawai, pushing beyond camp to a fangirl fantasy of irresponsible sweetness. This candy-colored city, with its pink blimps and anime posters, spills over with tweens, teens, and twentysomethings shopping in malls, flirting at stalls, and sipping bubble tea.

     

In the process, old stuff becomes cool.  Tradition, in the form of calligraphy and handmade paper, is a retro decorator choice, while letting your date clean your ears old-style makes him a friend with benefits.

     

It might all seem sappy, but like Tati’s Play Time and Wong Kar-wai’s Chungking Express, the film seeks to distill authentic poignancy out of kitsch, schlock, consumer clichés, lethal cuteness, and the detritus of urban lives. Frivolity must be good for something; why else would God give us giggles?  Comic form, as Shakespeare acknowledged, redeems a lot of silliness, especially if the gags are hurled at us with the ruthless conviction that anything goes.

Did I mention that all the roles are played by female performers?

In a switcheroo on Elizabethan theatre, globalization inverts the Globe. The film, a final title tells us, is dedicated to Shakespeare “but also to the patriarchy who would not allow female actors upon the stage.” The frisson is akin to that of Tsui Hark’s Peking Opera Blues and The East Is Red, where gender-blurring yields both humor and genuine feeling. From instant to instant, you see a character go male or female or something in between; a painted-on mustache and a swaggering gait become cosplay, not deep definitions of you. Unless you want it to be.

     

Identities are fluid. Okay, says Orlando, so he falls in love with Rosalind, then Roosevelt, and refinds him/her as Rose. What’s in a name? You get to call yourself, and be, what you wish, and love whoever.

     

Not a fresh-minted message these days, but the sparkle comes with how it’s all carried off. Every scene finds a clever way to amuse or bemuse. When Rosalind as Roosevelt slips into a trim suit, she pads out the crotch with a towel, and teeny gull-like waves waft out. That’s soft power, the equivalent of a mystic ring. Eventually she has to go along when Orlando visits the men’s room. While he stands at the pissoir, she ducks into a toilet pretending to take a dump, her groans covering the sound of peeling open a maxi-pad.

     

The project was co-directed. Wei Ying-chuan, a graduate of NYU’s Educational Theatre division, is a founder of Shakespeare’s Wild Sisters Group in Taiwan. Chen Hung-yi’s feature The Last Painting was chosen for IFFR in 2017 and won a best feature award at Cines del Sur. The pair bring an unflagging energy to the task of creating a paradise of easy living and loving–bereft of villains, open to any piece of harmless fun and heartbreak. As We Like It is a must for every LBGTQ film event, but its hella dirty fun for any festival whatsoever. Couples welcome.

Again, the IFFR provides a fine discussion of the film with the directors, moderated by our old friend Shelly Kraicer.

 

St. Tropez, mon amour

Suzanna Andler (2020).

Eric Bentley once described great serious literature as “soap opera plus.” Anna Karenina, Othello, and the rest offer us tormented love affairs, sexual jealousy, hidden schemes, and forced confessions of betrayal, but it’s all endowed with wider significance through characterization, implication, style. But can we have soap opera minus?

In Daisy Kenyon, Anatomy of a Murder, and other films, Preminger moves in this direction, banking the fires of conflicts drawn from lurid bestsellers, but other filmmakers have gone further in de-dramatizing melodrama. Dreyer’s Gertrud and some of Oliveira’s adaptations offer examples. Here we have the classic fraught situations, but muffled and fragmented and punctured through long pauses and wayward, looping, maddeningly banal conversations.

Marguerite Duras made this artistic strategy peculiarly her own, notably in her masterpiece India Song (1974) and its counterpart Son nom de Venise en Calcutta désert (1976). In a curious reversal, she often prepared the film first and then published the text as a quasi-play, as if scraping away the luscious imagery and ripe sound would create something even purer, soap opera distilled to Racinian starkness.

Suzanna Andler, a Duras theatre piece from 1968, has now been adapted to film featuring Charlotte Gainsbourg and three other players. The result isn’t as severe as the play reads, since director Benoît Jacquot has filmed it in a gorgeous villa overlooking the Mediterranean. It remains, however, in the tradition of kammerspiel. The bulk of the action takes place in a salon and the terrace outside, with one sequence, also in the play, set on a rocky beach.

The situation is sheer bourgeois melodrama. Suzanna is in a loveless marriage with the philandering Jean. She has apparently stayed with him for the sake of their children and the wealthy life they lead. Now Michel, a young journalist, has tempted her into a love affair, and she has for the first time cheated on Jean–who seems okay with it. Today’s crisis, if this counts as one, is her need to decide: Will she lease this villa for the summer with the kids? Or will she accept Michel’s invitation to go to Cannes? In the course of about four hours, she may make up her mind.

If some of my synopsis seems hazy, it’s partly because the exposition comes out in bits as Suzanna and others chat about her past, and partly because what she says may not be wholly truthful. She sometimes admits to lying. And what was her relation to the never-seen Bernard Fontaine, who has just been killed in a car crash? The blurry backstory is one strategy Duras uses to tamp down the melodrama, which usually gives its plots clear-cut contours and definite revelations.

In filming the play, Jacquot has taken an approach that approximates the rigor of Duras’s aesthetic. He has shot the blocks of action using slightly different techniques. Not for him obvious alternatives like color/ black-and-white or a range of tonalities. The differences are made harder to spot because Jacquot has not given us separate chapters corresponding to the act divisions; the scenes blend, punctuated only by long shots. So there are stylistic spoilers coming up.

At the start, Suzanna is shown the house by the real estate agent de Rivière. This segment is filmed in distant shots that emphasize the landscape and straight-on views of the sitting room opening out onto the terrace and the sea. The agent is seen from behind or at a distance, while the few close views we get concentrate on Susanna.

     

Staying behind alone, she falls asleep and awakes when Michel enters. This is a second phase of the play’s first act, and now Jacquot’s camera setups take a more oblique view of the room. The full-length windows dominate again, but now at an angle that recalls the magic mirror of India Song (on which Jacquot was an assistant).

     

The couple is often seen at a distance, but now closer views of Suzanna emphasize the mirror motif.

     

At a high point, the camera celebrates a momentary reconciliation with a track in to their embrace (the first such florid move in the film, I think).

In the sequence corresponding to the second act, Suzanna meets her friend (and Jean’s ex-lover) Monique. On the beach they talk of their pasts. Now the conversation is rendered in many rapid, tight shots of the two women. The orthodox shot/ reverse shot setups are sometimes given a strange emphasis when instead of A/B alternation we get two (but only two) variants of a view of each one as she speaks (A1/A2, then B1/B2). So a cut like this::

     

. . . is followed by ones like these:

     

Back at the villa, Suzanna answers a phone call from Jean, and they discuss their plans, with the uncertainty typical of all the film’s conversations. This scene is handled in circular tracking shots around Suzanna, from a moderately close distance.

     

As the conversation ends, Michel returns. After he reveals some key information about his relation to Jean, he stretches out on the sofa. In a long take running several minutes, the camera swings around them in a half-circle, clockwise and counterclockwise, often adjusting to her shifts in position.

     

The changing angle also captures Suzanna perched against a painting of very 60s boomerang shapes that echo the camera’s trajectory.

As the action approaches what might be a climax, Michel drifts out to the terrace and sits on the balustrade above the sea. Suzanna approaches.

Telling you what happens next would truly be a spoiler. On seeing it, I thought it was something that Jacquot added to the play, but nope . . . it’s there in the text, and he’s perfectly faithful to it.

As if all this patterning doesn’t look finicky enough, the scene on the beach is punctuated by a single shot of the Quai de Passy with a Métro train rumbling by.

This bump comes exactly halfway through the film, at the moment Suzanna mentions the surge of attraction she felt when Michel looked at her on their first encounter. Believe it or not, the line in the play also comes midway through the text. This alien shot functions expressively, I think. It underscores the epiphany Susanna felt upon learning she might be loved. Another filmmaker might have stressed the moment with music under her monologue, but Jacquot goes for a formal bonus: breaking the visual texture just here further articulates the design of his film.

The rigorous geometry Jacquot has clamped down on the play is interesting in itself, and the abstract array of options adds, I think, to the hieratic quality Duras is after. Yet each style matches the tenor of the action it carries and doesn’t conceal the subdued feelings rippling through the scenes. This dimension depends on Gainsbourg–her slim silhouette, her microdress, and especially her face, with her alert chin and hard mouth. Her vacillations have nothing of the diva about them, but still she stands forth as a new avatar of The Confused Woman so beloved of art cinema (Voyage to Italy, L’Avventura, The Headless Woman). Without those closer shots, the film might fall flat.

Once asked what would be his ideal final shot, Jacquot replied: “A distinctive glance [un certain regard] in close-up.” His ending delivers that.

Duras is doing something similar to what Wei and Chen do in As We Like It. She is seeking genuine emotion in clichés (unfaithful husband, wrung-out wife, surly rescuer). But she  hasn’t exempted her characters from social critique. Hiroshima, mon amour renders the meeting of two lovers as an intertwining of two national histories. The colonialists of India Song, drifting through their sparsely attended embassy parties, trying to replicate salon society in the tropics, cannot hear the voices offscreen of the people they subjugate. Likewise, Suzanna’s anxiety may or may not register some distant tremors. In summer of 1968 her world is sliding into something she isn’t prepared for. Far away from St. Tropez, in Paris students are hoping to find their own beach, but they’re doing it by tearing up the pavement.


Again, thanks to Gerwin Tamsma, Monika Hyatt, Frédérique Nijman, and their colleagues at the International Film Festival Rotterdam for allowing us to visit their event virtually. Here’s to another fifty years of ambitious programming!

A very helpful edition of Suzanna Andler has been published in conjunction with the film’s release. It contains a lot of stimulating background information and critical commentary. Bentley’s comment about “soap opera plus” comes from The Life of the Drama (Applause, 1991), 14. Thanks to Kelley Conway for sharing with me the Jacquot interview  in “Réponses à tout,” Libération (14 May 2004), 1.

Jacqout’s rendition of Duras’s play exemplifies what I called in Narration in the Fiction Film “parametric narration.” This rare approach consists of playing out a range of expressive possibilities, scene by scene, in ways that both shape the ongoing plot and “anthologize” sharply contrasting cinematic techniques. Noël Burch first proposed this idea in his enduring Theory of Film Practice (1973), although Eisenstein and Bazin envisioned it. But then they envisioned everything.

P. S. later: The Rotterdam prize winners have been announced (per Variety).

As We Like It (2021).

David Bordwell
top of page

have comments about the state of this website? go here