Casablanca (1943).

DB here:

How could I have written a longish book on 1940s Hollywood and have devoted so little space to Casablanca?

This question was brought home to me when Pauline Lampert, mastermind of Flixwise, asked me to be a guest on her podcast….and to talk about Casablanca. You can listen to our conversation here. Yes, Joan Crawford is involved.

Reinventing Hollywood mentions Casablanca in a few places, but it doesn’t discuss it in the depth it devotes to, say, A Letter to Three Wives or Cover Girl or Five Graves to Cairo or Unfaithfully Yours, let alone Swell Guy or Repeat Performance or The Guilt of Janet Ames. I suppose it’s partly because many of those movies are less famous and more peculiar.

In addition, I confess that I’ve never been a big fan of Casablanca. I admire it as a solid piece of work, but it hasn’t aroused my passion. I’m much more emotionally attached to How Green Was My Valley, The Magnificent Ambersons, Shadow of a Doubt, Notorious, Meet Me in St. Louis, Mildred Pierce, The Little Foxes, On the Town, and many other movies of its time.

Admittedly, my book doesn’t dwell on most of these either, although I’ve written about most of them elsewhere (including on this site). Basically, I suppose I neglected Warners’ evergreen classic because its canonical status made it unnecessary for me to talk about it. It’s a typical 40s film that everybody knows well, and I guess I trusted that interested readers would apply to it the ideas I float in the book. And another factor may have blocked my considering Casablanca.

 

Learning from a classic

When historians want to explain changes in film artistry, they have some options. One possibility is to focus on the influential individual, the great artist who inspires successors. Charles Rosen makes this case in The Classical Style, which concentrates on Haydn, Mozart, and Beethoven as prime innovators. Traditional film historians have singled out Griffith for this role; I’d add Ozu and Antonioni, in their respective contexts, as significant influencers.

A second option is to zero in on an individual work that became a prototype for other works. The Birth of a Nation, Potemkin, Citizen Kane, and The Bicycle Thieves have played this role in film histories.

A third possibility is to explain change through patterns of collective practice. Creative choices, disclosed in several (perhaps minor) works, propagate through a community and eventually establish norms. This seems to have been what happened with tableau staging and the system of continuity editing. They emerged in bits and pieces, as ad hoc practices that were forged into firm craft practice.

Without denying that some individuals, like Welles and Hitchcock, matter, and without denying that there are powerful models, such as Kane, my book was mostly committed to showing how narrative norms, pressured by competition and other factors, mutate across the 40s ecosystem. This emphasis on norms is one aspect of my research program.

So in re-watching Casablanca for my meeting with Lady P.’s podcast, I noticed some norm-abiding and norm-tweaking things I could have written about.

Classical plot construction (of course). Despite the legend of last-minute screenplay fixes, this is a tightly organized plot. Kristin’ s four-part structural template is in force. After shrewdly distributed exposition, Laszlo and Ilsa stroll into Rick’s at the 25-minute point. At the midpoint (about 50:00), Rick calls Ilsa a whore and she departs to meet Renault. The climax, I’d say, starts around 80:00, when Rick and Ilsa reconcile and he hatches his plan to rescue her and Laszlo.

The parallels are also well-carpentered. Rick is compared to Laszlo; both have been activists. Rick’s also like Renault, in that both are now in-between men, apolitical cynics who will convert to the cause. Ilsa is paralleled to Rick’s high-strung paramour Yvonne, as well as the Bulgarian woman who has concealed from her husband the fact that she traded her body for safe passage out of Casablanca.

Ticking the 40s boxes: Voice-over narration to open the thing? Check. Crisis structure that launches an explanatory flashback? Check.

Franker sexuality and Parker Tyler’s Morality of the Single Instance? Check-plus. When Ilsa comes to Rick’s apartment, they obviously do the thing. He later says they “got Paris back last night.”

And of course it’s a feast of 40s character actors, including Rains, Lorre, Greenstreet (in a silly fez), Veidt, Kinskey, Dalio, and Sakall (aka Cuddles).

     

Even John Qualen shows up. Casablanca is as much an oddball hangout as Casablanca is.

Domesticating deep focus: Part of the influence of Citizen Kane stemmed from its flaunting of big-foreground depth. A head or an object would be pasted in the front plane, and other figures would stretch out in the distance.

     

Kane wasn’t the first film to use this strategy, but Welles and Gregg Toland made it vivid by playing out these shots in very long takes. After Kane, 1941 became the Year of the Big Head, apparent in The Little Foxes and, surprisingly, Ball of Fire. Toland shot these.

     

But these films didn’t treat such compositions in long takes. Wyler and Hawks integrated the wide-angle shots into a standard analytical editing breakdown, using the big foregrounds for orthodox shot/ reverse shot. We see the trend as well in tighter-than-usual over-the-shoulder shots like these from The Maltese Falcon, also 1941.

     

Curtiz wasn’t as flamboyant a stylist as Welles, but he often added pictorial zest to his shots. He too integrates the wide-angled foreground into standardized setups and editing patterns. Interestingly, his cinematographer was Arthur Edeson, who shot Falcon.

     

     

In this respect, Casablanca exemplifies how a new norm was diffused as a toned-down version of a more self-conscious technique.

Smooth visual and verbal narration: This time around I admired the opening passage, with rapid exposition based on adjacency. We start at street level, as the police capture suspected agents of the Free French. One is gunned down, in a moderate deep-space composition. And everything is explained by a chatty pickpocket.

     

The Bulgarian couple is planted in a panning shot that follows the secret agent and settles on them. (Again, the wide angle accentuates foreground action.)

               

We get continuity by contiguity. A neat touch: Major Strasser’s plane arriving is glimpsed in conjunction with Rick’s sign, setting up that location well before we visit it. (See still up top.) At the airport, Renault closes the loop, explaining to Strasser that everybody comes to Rick’s. That initiates a dialogue hook to the café, where we’ll spend the next thirty minutes of screen time. And the arrival of Strasser’s plane rhymes neatly with the departure of the one carrying Ilsa and Laszlo at the end.

Motifs, motifs: In a movie where every scene has a famous tagline, I was struck this time by one tweak. Meeting Strasser, Renault says that he’ll round up “twice the usual number of suspects.” In the epilogue, that phrase becomes the nonchalant one everybody remembers, “Round up the usual suspects.” It’s a reminder of corrupt policing, a guarantee that Rick will not be charged, a sign of Renault’s new loyalties, and a contrast to Renault’s initial obedience to the man Rick has just killed.

Seeing eye to eye: I’ve written elsewhere (here and here and here) about the importance of eye behavior in storytelling cinema. In Casablanca, I noticed how Edeson’s cinematography made the women’s eyes glisten. This is standard female lighting for the period, but Ingrid Bergman often gets an extra princess sparkle.

     

As for Bogart, often a blinker, he lowers his eyes ambiguously when he says that Renault has always kept his word.

When the Bulgarian woman asks about a woman who keeps secret something to save her husband, he manages to shift into a stricken stare for eleven seconds: “Nobody ever loved me that much.”

          

Actually, Bergman beats him when Sam plays “As Time Goes By.” Ilsa stares, slightly shifting her head but with a largely blank expression, though her mouth opens slightly.

     

She doesn’t blink for twenty-four seconds. Nearly all the action is in those eyes, as a glint of light jumps from her right eye to her left.

     

A 1940s movie, over-solemn as it can sometimes be, has this virtue: it can show people thinking.

 

When you try to study filmmaking norms, often you find that the individual film becomes a “tutor text,” as the French Structuralists used to call it. The movie shows you things about movies. His Girl Friday, I’ve argued, is such a film for me. Maybe Casablanca should be. In any event, we can learn a lot from it, and in the process still enjoy its drama of love sacrificed to political commitment.

---

Thanks to Pauline for having me on Flixwise. Kristin’s analysis of four-part plotting, first floated in Storytelling in the New Hollywood, has been applied and expanded in my The Way Hollywood Tells It and many of our site entries, as here and here. I discuss Parker Tyler’s Morality of the Single Instance in The Rhapsodes. My deep-focus arguments can be found in Chapter 27 of The Classical Hollywood Cinema and developed at greater length in On the History of Film Style, soon to be available in an updated version on this site. See also this entry and this one and this one.

Casablanca.

La La Land.

DB here:

Between the end of principal photography on First Man and the start of post-production, Damien Chazelle squeezed in a visit to the UW–Madison. We’re very glad he did. A hell of a time was had by all.

His visit culminated a Cinematheque series devoted to his work. On Friday 23 February we picked him up at O’Hare and had a fine ride back talking about film and less important things. Then he visited our archives at the Wisconsin Center for Film and Theater Research; on the right, he examines an original Final Revised script of Citizen Kane.

After that, he sat down for a conversation about his career with a hundred or so students. At a quick dinner, he and our Cinematheque impresario Jim Healy gave dueling impersonations of Michael Gazzo as Frankie Pentangeli. Damien then plunged into a long Q & A with a full house who had just seen La La Land in 35mm.

Next morning he met with Criterionistas Kim Hendrickson and Grant Delin for a FilmStruck segment. Then, in a discussion with Kelley Conway, he introduced a string of films he curated for the Cinematheque. But he wasn’t off the hook, because driving back to O’Hare with Kelley and Jeff Smith, he was immersed in more film talk.

Damien proved himself the ultimate guest—friendly and generous, enthusiastic and excited, free of airs and snark. We learned a lot from him. Herewith, a sample.

 

A Direct-Cinema musical

Guy and Madeline on a Park Bench.

Although he considered a musical career, film was Damien’s first love. He wrote scripts in middle school, transcribed movie dialogue from VHS tapes, and as an undergrad watched films in Harvard’s magnificent archive. The film program there, with leaders like Alfred Guzzetti and Ross McAlwee, stressed documentary and experimental film, and the exposure stuck. Among the films Damien curated for our Cinematheque show were the Rouch-Morin investigation Chronicle of a Summer and Su Friedrich’s Sink or Swim.

No surprise, then, that his first feature, Guy and Madeline on a Park Bench, was shot in a Direct Cinema mode. It’s got light leaks and run-and-gun footage, complete with bumpy handheld pans and zooms. To get around problems of inexperienced actors, Damien told some of them that it was a documentary. The 16mm project was produced over three years; sometimes the exposed films sat in the lab while Damien drummed up donations from friends, family, and strangers. (Writing blind to Harvard alums, Damien got a donation from John Lithgow.) When a processing accident ruined some footage, Damien’s producer talked the lab into free work for a time.

Guy and Madeline cuts among three characters: trumpeter Guy, his ex-girlfriend Madeline, and his new girlfriend Ilena. Like a Nouvelle Vague film, it relies on chance encounters. Madeline is emotionally wrenched by the breakup with Guy, and we follow her efforts to find work and a new partner. Ilena’s semi-reluctant meeting with an older man who brings her home to meet his daughter reminded me of the moment in Shoot the Piano Player when Charlie, running from the thugs, falls into step beside a stranger who tells him his own troubles. And of course the title characters recall the separated lovers of The Umbrellas of Cherbourg.

For all its documentary textures, the film at times becomes what J. Hoberman called a Mumblecore musical. But there’s a gradual shift to the full-blown show-biz mode. Damien talked of the thrilling moment in Hollywood musicals when realistic presentation of a scene gives way to nondiegetic music and the characters leap to a new, more ethereal level. Guy and Madeline presents this transition in gradual doses.

At first, the numbers are motivated realistically. Guy, an African American, plays trumpet with a jazz ensemble, so we get scenes of their performance at a local hangout. We move a bit further toward stylization with a party sequence that induces some talented kids to indulge in singing and tap-dancing among their friends—captured in casually imperfect framings.

The transition to pure musical fantasy comes forward after the breakup in a solo number, with Madeline singing a soliloquy as she wanders in the park. There’s no sense of an audience; this is a private reverie. (A whiff of this tune, heard on a car radio, makes its way into La La Land‘s opening shot.)

When Madeline takes up work at a diner, the components come together in an all-out production number addressed to us.

In an echo of Bande à part’s “Madison” sequence (lucky name), shooting a dance number in cinéma-vérité mode brings out an intriguing friction. It’s the same kind of productive clash we get on the soundtrack, between Justin Hurwitz’s shimmering Legrand-inflected score and varieties of jazz (Dixieland, Guy’s cool composition for Madeline). And like Nouvelle Vague characters, these people are devoted to books, the arts, and self-exploration.

As a “staged documentary” Chazelle’s film parallels Chronicle of a Summer in an intriguing way. That film starts as pure Direct Cinema, with investigators stopping people on the street to ask them questions. But as we get to know the group the film concentrates on, there’s a lot more control and “fictionalization.” There are precise matches on action, for instance, with camera ubiquity indicating careful restaging.

     

This rigging doesn’t damage the film as a document of summer 1960. Damien learned from it that you can make a truthful movie by “creating a situation with less and less acting to do.” Given this hybrid quality, Chronicle of a Summer becomes a vivid example of a moment when a film mode is “figuring itself out.” Its self-conscious artifice, which includes participants watching themselves during a screening, was foundational for the New Wave. “You watch a language being born.” That language was also political, as Damien pointed out: The film summons up memories of the Holocaust and glimpses of the Algerian war.

In other respects, Damien’s first film looks forward to La La Land thematically and formally. Guy and Madeline starts with the moment of the couple’s breakup (on the bench) and flashes back to vignettes of their love affair before returning to the bench. This opening loop is like the one that jumps from Mia’s night out back to the traffic jam and then follows Sebastian. A large stretch of each film’s plot is about how the couple’s lives converge and diverge.

Similarly, when the signature tune “I Left My Heart in Cincinnati” is played, shots of Madeline and Guy frame a flashback to the combo’s earlier performance, as if they’re sharing the memory. Something similar happens at the climax of La La Land, in what seems to be a mutual vision of Sebastian and Mia’s alternative future. As often happens in Chazelle’s cinema, epiphanies burst out in moments of musical performance.

 

Blood, sweat, and tears on the drumhead

Grand Piano.

Despite playing many festivals and winning critical praise, Guy and Madeline didn’t open any doors in Hollywood. Damien picked up odd jobs, not all film-related, while writing commercial genre screenplays. He sold a kidnapping script (not made) and Grand Piano (2013), skillfully directed by Eugenio Mira. He began getting assignments like The Last Exorcism Part II (2013) and he contributed to the screenplay for what became  10 Cloverfield Lane (2017), released long after he’d worked on it.

I found Grand Piano pretty impressive on the big screen. Chazelle’s script and Mira’s direction create a solid thriller built around the situation Hitchcock designed for his versions of The Man Who Knew Too Much. Most of the action takes place during a concert celebrating the return of a traumatized pianist to the stage. As he’s about to start the program, a sniper uses cellphone messages and scribbles on the score to demand a perfect performance of the florid piece that spooked the pianist years before.

At first restricted to the pianist, the film’s viewpoint widens gradually to include others, and soon crosscutting builds tension. The tormenting voice (“Play one wrong note and you die”) calls to mind the music teacher in Whiplash. As in a classic thriller, the climax arrives when the victim must fight back. And as in Whiplash, the performer wins using the only weapon he has: nearly crazed virtuosity.

Damien now thinks that the long germination of the scripts for Whiplash and La La Land made them better. As financing kept falling through, the films gained more layers. Whiplash (2014) found a home first, with Blumhouse producing and helping with the financing. It was their idea to shoot a scene  to show investors (we screened it in our series), and the project found financing at Bold Films.

Given a $3 million budget and a 20-day schedule, Whiplash demanded meticulous storyboarding and very little coverage. Like Hitchcock and Leone, Damien shot only what he needed. He used two cameras for the rehearsal scenes and three for the climactic concert. The cuts and camera moves were planned to coincide with measures of the music.

Damien calls Whiplash a film about music (the same could apply to Grand Piano). It owes a lot to the sports-film genre as well; Damien envisioned its punishing force as indebted to Raging Bull. He turns big-band drumming into blunt-force trauma, with gory drumheads and cymbals. Sam Fuller would have approved.

Like Guy and Madeline and Grand Piano, Whiplash culminates in a musical performance that carries a powerful emotional impact. No wonder that as a kid Chazelle studied one-reel movies of classic drummers, then started to think of the shorts as films in their own right. In this spirit he curated for us two Dudley Murphy shorts, St. Louis Blues (1929, with Bessie Smith) and Black & Tan (1929, with Duke Ellington), along with the 1954 documentary Jazz Dance, a night on the town that explodes with pure human happiness. In all these, music-making is pushed to the edge of ecstasy.

This time around with Whiplash (good name for a movie about sadomasochistic musicians), I noticed its straightforward classical construction. Damien says that he learned screenplay construction after moving to LA. Its tale of a boy caught between a good but weak father and a punishing, strong one gains strength and sharpness from its traditional four-part plot.

At the crucial 25-minute mark, Fletcher wins Andrew’s trust. Four minutes later, in the performance of “Whiplash,” Fletcher is bellowing and Andrew is sobbing. First reversal noted. The second part, the Complicating Action, interweaves Andrew’s romance with Nicole, his persistence in drumming, and his fraught relation with his family. This part culminates at the midpoint with Fletcher’s giving Andrew a new rival, which impels Andrew to break up with Nicole. In the Development section, Andrew suffers more setbacks. A harrowing car accident leads him to botch a major competition and assault Fletcher. He leaves school, accuses Fletcher of abuse, and abandons drumming.

After he discovers Fletcher playing piano in a club, he agrees to join his new combo, which preciptates the climax: a competition performance at which Andrew, realizing that Fletcher is out for revenge, seizes control. The result is another burst of barely controlled frenzy, complete with unmotivated bursts of light spattering Andrew in the last shot.

     

Whiplash is a film without pity. Andrew’s rejection of Nicole suggests that he’s become obsessive, and after his scuffle with Fletcher he’s drained and numb. And no sympathy is extended to the monstrous Fletcher. Damien avoided what he called the “rubber ducky” moment that shows this man to be damaged by some childhood trauma. We get no explanation of his ruthless brutality; he’s simply a force to be fled or fought. (Damien told us that he modeled Fletcher on a music teacher he’d had; the original probably wasn’t as nasty, but Damien wanted the film to convey how frightening he was to a fifteen-year-old.)

At the end, Andrew earns a glint of triumph, but the reverse shot shrewdly withholds from us the expression that might warm us up to this man. His sliced-off smile and slight nod are all it takes for Andrew to react.

     

Still, his grudging approval means that Andrew has won over one scary dad.

 

Embarrassing yourself and your characters

At 29, Chazelle found himself with a hit, confirming the Magic Number 30 Rule. Whiplash made a splash at the Sundance Film Festival and went on to be nominated for several Oscars, winning three. It also brought in a lot more than its cost. Damien could now reignite La La Land.

Lionsgate, via Summit, picked it up and production began. There were 40 days of shooting across 65 locations. The production was able to be so efficient because of careful planning and the reliance on long takes. “The long take has become fashionable,” Damien says, who knows his film history: “But it’s actually the most old-fashioned kind of thing.” Whiplash is an editing-driven movie, but La La Land relies on many fewer shots. The moments of shot/reverse shot–notably in the spoiled dinner when Sebastian is briefly between tour gigs–gain a prominence they don’t have in most movies.

In shooting, the morning was given over to rehearsals, followed by a great many takes–often required to sync the actors to playback. Around about take twelve, Damien recalls, things started to crystallize, but sometimes as many as twenty takes were needed. While Whiplash stayed tight to the screenplay, La La Land was heavily improvised. The visuals were pre-designed, but the relationship at the center needed a casual feel, as if the characters were tossing off their lines.

Damien had thought he would simply be able to cut together the “one-ers,” but editing took five months, not least because he and his long-time editor Tom Cross played with many versions. Every number was a candidate for deletion, including the freeway opening. (Yeah, I shuddered.) There was also some digital adjustment of the 35mm original. David Koepp wrote me:

Maybe ask Chazelle about how beautifully he used color to direct our eye in his LA LA LAND opening. Always the right burst of costume color directing us to the right spot at the right time… although I guess that’s more of a compliment than a question.

Damien explained that he made those colors pop a bit more by digitally toning down other costumes in that intricate opening sequence.

La La Land has steadily grown in my regard and affection; I think it’s one of the best recent American movies. Just the title gets you going. “La la” suggests music, but also the self-absorption of jamming your ears lalalala. LA is a town of airheads, but it can become a town of worthwhile fantasy too. Damien spoke of most movies trying to make fake sets look real; he wanted to “take real stuff and make it look false.”

This time around, I was struck by the film’s harsh side. It’s pretty hard on mainstream Hollywood, from the smug partygoer who says he’s really good at world-building to Mia’s superficial roommates. Their anthem “Someone in the Crowd” is about careerism, but it becomes for her about the search for a soulmate.

Then there’s Sebastian. A lot of Hollywood plots work only if the guy is a jerk. In Whiplash, Andrew turns smug when he thinks he’s Fletcher’s pet, and he dumps Nicole heartlessly. (He’s becoming a bit like Fletcher.) In La La Land, I began to see Sebastian as a stubborn nerd, refusing to play the cocktail-bar set list and ranting about jazz to anyone who’ll listen. Ryan Gosling’s ingratiating performance makes this nerd more likable, but as written the character is pretty arrogant.

One scene that puzzled me now makes sense in a larger pattern of Seb’s obtuse, evasive behavior. After he learns he may be kept from attending Mia’s show, why doesn’t he phone her? We see him brooding outside the music studio.

He may think he can still make it in time, which would reflect his somewhat risky self-assurance. But Damien pointed out that elsewhere in the film he’s not seen using a cellphone, or for that matter a computer. Old school as he is, he seems wary of modern technology. He drives an utterly impractical Buick convertible. He plays cassette tapes and LPs and his apartment’s phone is an old-style handset, antenna and all. An omitted screenplay scene showed him in a movie audience ranting at somebody using a phone, thereby disturbing the viewers more than the caller has.

I’m being too hard on Sebastian, of course. We admire his idealism, his tenacity, and his romantic attachment to what he thinks is the best of the past. Still, Damien has remarked that he sees sides of himself in both Andrew and Sebastian, which reminds us that “commercial” films can also be personal ones. For him, the strongest creative choices risk exposing you. “If you’re not embarrassing yourself, you’re not doing your job.”

If Sebastian is too willful, Mia is too eager and desperate. “I can do it differently,” she tells the audition staff after they’ve brushed her off. Sebastian and Mia complement each other. His cockiness (“Fuck ’em”) pushes her to mount her one-woman show, while she tries to steer him back to his basic commitments. The larger theme seems to me that the most vital art comes from yourself, be it your memory of a Francophile aunt or your irrational attachment to classic jazz. Instead of having to fit into prefab TV characters, Mia gets her breakout role in a film that will build its script around her personality.

Damien spoke of the musical as Hollywood’s most avant-garde genre. That partly stems from the transition from realism to fantasy that launches a number. This shift provides the film’s final turning point, with Mia’s audition; for once a Chazelle film makes its musical climax a subdued one, but it’s no less a demonstration of the performer’s authentic emotion. Art’s power comes from novelty (“new colors to sing”) grounded in sincerity and self-awareness, even if by some standards it seems awkward and geekish.

The avant-garde overtones are also a matter of how musicals make real locations look unreal—as Demy films memorably show. So it was uncannily appropriate that Damien asked us to introduce La La Land with Bruce Baillie’s All My Life (1966). A slow pan left across a fence and flowers gives way to a diagonal tilt up to the sky; the whole accompanied by Ella Fitzgerald and Teddy Wilson.

     

La La Land might well be a sequel, as we tilt down from another blue sky to a gridlocked freeway.

     

As Baillie turns a prosaic bush and fence into an audiovisual flow, so the opening of Chazelle’s film takes the banality of a traffic jam and makes it an explosion of youthful hope and energy, complete with somersaults.

The sheer cinematic exuberance of La La Land will, I think, keep the film alive for a long time. “Every scene, a new idea”: Damien quoted Arnaud Desplechin quoting Truffaut. Many parts of La La Land put nifty tweaks on the conventions of comedy, drama, and the musical. There’s the “enacted” slow-mo at the party, the iris around a kiss, and the montage rendered as a flash-forward from a duet at the piano (“City of Stars”). There’s often a tweak on what might have been perfunctory filler. The exit-on-an-elevator shot is lit and costumed so as to (a) suggest the conformity of the dress code for an audition; (b) emphasize the height of her rivals; and (c) accentuate the spill on the less glamorous Mia’s blouse. Her disadvantages are diagrammed.

Then there’s the idea of having a “real” dream ballet at the planetarium and a virtual one at the end. Speaking of the end, I especially liked the head-fake at the start of the present-time part. By showing Mia on the Warners lot and Sebastian in his club, we’re invited to infer for a moment that they stayed a couple, before revealing that she’s actually married to an easygoing beefcake and Seb still lives alone.

Pitching La La Land, Damien found that many producers insisted that the couple unite for a happy ending. Damien objected that many of the great romantic films, including Casablanca, A Star Is Born, and Gone with the Wind, center on lost love. Still, he found a way to a happy ending by offering an alternative outcome that many viewers will prefer.

True, it’s sad. But Jacques Demy once remarked that sad movies make him happy. For me, La La Land is that sort of movie.

 

How much does cinephilia help a director? I’d expected Damien to recommend the sort of movie immersion he had as a kid. And he admitted the power of the past. “I can’t unwatch the movies I’ve seen.” But some great directors aren’t cinephiles, he granted. He cited Bresson and Dreyer; I thought of Ford. What’s important, he suggested, is a relation to some, any art form–if not film, then visual arts or theatre or literature.

Maybe the best of both worlds is to be a young filmmaker who knows both film and another medium, such as music, and thinks as an audiovisual artist. Damnien remarked that in writing he starts with images rather than words but then lets the dialogue focus the scene. Interestingly, Eisenstein taught his students to stage a scene first as if it were in a silent film, then revise it with music, color, and (only then) dialogue. That assured that pictorial storytelling would be foremost.

Kristin and I were gratified to hear that Damien has over the years read several things we’ve written. In turn, he taught us a lot. His visit reminded me that one path to filmmaking achievement is just thinking about your craft and your choices, in light of your life experiences and your encounters with powerful art. He passed that lesson along to the hundreds of people who came to learn from him.

---

Thanks to Damien Chazelle and Alissa Goldberg for making the visit possible. Thanks as well to J. J. Murphy, Mike King, Ben Reiser, Matt St. John, Mary Huelsbeck, Amy Sloper, Maria Belodubrovskaya, Erik Gunneson, Jason Quist, Kim Hendrickson, and Grant Delin. Event planners Kelley Conway, Jeff Smith, and above all Jim Healy, Cinematheque Director, deserve massive gratitude as well.

We have other discussion of La La Land on this site: my search for some of its roots in 1940s innovations, and my analysis of its song plot. There’s also a wide-ranging conversation among experts Kelley Conway, Eric Dienstfrey, and Amanda McQueen. Jeff Smith weighed in on the film’s score, correctly predicting its Oscar triumph.

P.S. 8 March 2018: Many thanks to Steve Elworth for a correction about All My Life.

Kelley Conway interviews Damien Chazelle.

 

DB here:

First, the latest installment of our Observations on Film Art series has dropped, as the kids (and now the grownups) say. It features Kristin on unconventional lighting (including darkness) in the great early French sound film Wooden Crosses (1932). This intense World War I drama boasts scenes of Fuller-like frenzy, mixed with somber passages. It’s by Raymond Bernard, a director who was a bit obscure for a while, but who gained great prominence with the rediscovery of his remarkable Miracle of the Wolves (1924).While that film doesn’t seem to be available on US video, several other  Bernard films are streaming on the Criterion Channel of FilmStruck.

The arrival of Kristin’s entry tallies with the new and expanded FilmStruck. To the 1200 or so features already in the channel’s library, TCM has added 600 classics from MGM, Warners, RKO, and other studios. Many of these titles, including Citizen Kane and The Thin Man, have never been available on streaming before.

The price remains the same: $6.99 per month for vanilla FilmStruck, $10.99 for it and the Criterion Channel (which nearly all subscribers take). You can get the whole package on a yearly basis for $99.00. Yes, I bought a subscription.

Third, but no less big a deal, we just learned that FilmStruck has launched in the UK as well.

The choice of titles is smaller, partly because some films are held by other licensees in Europe, but there’s still a vast array. Cost is again very reasonable:  £5.99 per month, £59.90 for a year and two free months. We think Observations installments will be available on the UK site.

I sometimes wonder how I’d have turned out if I’d had so wide and deep an access to films during the 60s. Would I have read books or listened to music? Maybe all that kept me balanced was limited access to movies. Then again, I spent a lot of time ferreting out finds–time that could have been spent watching them. I grew up in a film culture devoted to seeking whispered-about rarities and traveling to see them. Kristin and I once went to a screening of L’Age d’or supplied by a protective collector who brought the nearly unseeable film with him on the plane. Now, in the wild spiral from scarcity to superabundance, all it takes to see Buñuel’s masterpiece is pressing your remote or clicking your trackpad and paying your credit-card bill.

Life is a trade-off, but still….pretty nice to have choices. Speaking of which, our earlier installments on the Criterion Channel are here.

---

As usual, thanks to the Criterion team: Peter Becker, Kim Hendrickson, Grant Delin, and their colleagues. They not only make nice movies available, they make nice movies about them.

Michael Koresky has a wide-ranging introduction to Bernard’s 1930s films on the Criterion site, and Phillip Lopate has a characteristically engaged appreciation at Cineaste.

Wooden Crosses.

DB here:

The Donovan Affair was Columbia’s first all-talking picture, and Frank Capra’s as well. We discussed it a little in an entry devoted to a Cinema Ritrovato retrospective of Capra’s films. The film lacked, and still lacks, a soundtrack, which may be lost.

The Donovan Affair is an unusually fluid early talkie, and that led me to speculate we might be seeing the silent release version. Wrong! It really does survive only as a sound film without a soundtrack; the Library of Congress print I studied last month has a leader labeled “synchronized version.”

The clunky plot didn’t improve on repeated viewing, but the film did teach me some things about those transitional years 1928-1932, when filmmakers were figuring out how to make a sound feature. I thought I’d share some of those findings with you today.

But I must warn you that there’s one big spoiler coming up. I tell you whodunit. It’s necessary to make a point, but I will warn you just before the offending paragraph, so you can skip if you wish.

 

Blood-spattered footlights

Murder ran wild on the Anglo-American stage of the 1920s. While melodramas of love and betrayal waned, mysteries rose in popularity. There were plays about gangsters, trials, and domestic homicide. There were comedies of lethal intrigue in spooky settings, like The Bat (1920), The Last Warning (1922), and The Cat and the Canary (1922). A little more seriously, in response the rise of the genteel British detective novel (Christie, Sayers, Allingham et al.), there emerged plays that dumped murder into society drawing rooms.

You know the format. The setting is typically a mansion, with portraits, plush salons, and an impressive library. The victim, usually a bounder, deserves killing. The suspects are both high and low: businessmen, doctors, lawyers, dowagers, flappers, playboys, ne’er-do-wells, gangsters, and servants. Into this ménage steps a police inspector, often with a bumbling assistant, who will more or less skillfully reveal the culprit. In the process, though, someone else is likely to die.

The prototype is Bayard Veiller’s play The Thirteenth Chair (1916), but by the 1920s Britain and the US produced a host of popular and well-reviewed drawing-room mysteries: The Nightcap (1920), In the Next Room (1923), The Creaking Chair (1924), Interference (1927), The Man at Six (1928), The Clutching Claw (1928), and The Canary Murder Case (1928). The most famous entry in this cycle is probably Patrick Hamilton’s Rope (1929), which stands out because it presents the crime from the killers’ viewpoint. The question isn’t “Who did it?” but “Will they get away with it?” Rope’s “inverted” structure (no pun intended; it’s a trade term) was anticipated by A. A. Milne’s The Fourth Wall (1928).

Owen Davis’s play The Donovan Affair (1926) fits snugly into this cycle. The action starts with Inspector Killian and his assistant Carney ushering the suspects into the Rankin family library. The rakish Jack Donovan has been murdered at a birthday dinner. The circumstances are odd: To demonstrate a glowing ring Donovan wears, the lights were turned out. In the darkness, he was stabbed with a carving knife.

Most of those at the table have good reason to kill Donovan. He was having affairs with married women and had an eye on Rankin’s daughter Jean, whom David hoped to marry. The first act is centered on exposition that emerges from Killian’s questioning of the guests. One, Horace Carter, comes forward to declare he knows who the killer is, because the victim’s missing ring was slipped into his pocket. Killian doubts that the ring could have been visible, and so the lights are switched off a second time as a test. Bad idea: When the lights come back on, Horace has been stabbed with the same knife. Score one for law and order. Later one character will shoot another as Killian looks on.

The next two acts expose more motives, lay false trails (David has brought a gun to dinner and has blood on his cuff), and string out physical clues, including a fake ring and some threatening letters to Donovan. In frustration, Killian leaves the suspects to pool their information, in hopes that they’ll discover the killer. Once more the lights are put out to allow the culprit to replace the ring and a missing letter. The killer is revealed, and there’s a struggle in semidarkness. Killian and his men burst in to seize the guilty party.

Among all this claptrap, the blackout gimmick is worked very hard. Bayard Veiller’s Thirteenth Chair had staged a stabbing in a dimly-lit séance, and Davis himself had let the electricity fail in The Haunted House (1924). The blackout-covering-a-killing would be reused in The Spider (1927). The Donovan Affair justifies the device through the luminous ring. In plot terms, it’s another red herring, but it plausibly motivates darkening the stage during the murders. The effect seems to have come off.

The audience found itself frequently strung almost to the endurance limit, and now and then a voice from the crowd would beg a relentless actor not to turn out those lights again. For it was generally in the darkness that stabbings, shootings and assorted mayhem would stalk through the play.

It must have been riveting to see the ring (a green-tinted flashlight) floating around the darkness. Capra’s film is at pains to reproduce this signature effect, but with some differences.

The screenplay subtracts some of the characters and adds a sinister caretaker and Porter, a gambler to whom Donovan owes money. The plot begins well before the crime, with an opening that establishes Porter among his cronies. A second scene shows Donovan brushing off Mary, the Rankin family maid; in the play their liaison is revealed very late. Then guests gather for the birthday dinner, and after some salon byplay they sit down to eat, with Donovan demonstrating the power of the cat’s-eye ring. In the darkness he’s murdered.

A title, “One hour later,” introduces the main stretch of the film, which corresponds more closely with the play’s action. Inspector Killian arrives and sizes up the suspects. Most of the play’s clues emerge, with suspicion scattered around freely.

Like the play, the film is built around blackouts showing off the wandering ring, but now these are motivated as reenactments of the crime. Killian puts Porter in Donovan’s seat, and in the darkness he becomes the second victim. After more intrigue, Killian demands another go-round, and this time the killer is exposed.

The film’s blackout scenes run 36 seconds, 42 seconds, and about 100 seconds for the climax. This last scene varies the treatment by showing some silhouettes.

     

These scenes feel surprisingly protracted: 30 seconds is a long time on the screen. Likely there would have been dialogue and sound effects, as there are in the play’s blackout intervals.

Here’s the paragraph with the spoiler. In both play and film, the butler Nelson is revealed as the murderer. (At this point in history, having hired help commit the crime wasn’t yet out of bounds.) Nelson is in love with Mary, whom Donovan has seduced. The film hints at this in a way that the play doesn’t. Mary is introduced early as Donovan’s mistress, Nelson seems to have eyes for her, and during a test to see if any woman’s handwriting matches that on a telltale letter, he lingers solicitously over Mary.

Variety praised Capra’s film as well-constructed and surprisingly funny, chiefly because of a whimsical old couple added to the ensemble. Capra claims that The Donovan Affair taught him to add doses of comedy as much as possible: “Comedy in all things.” Actually, though, Owen Davis claimed that he was teasing the genre in his original. “The Donovan Affair was, naturally, a success, although I am, I think, the only one who knows that it was as deliberate a burlesque as The Haunted House had been.”

 

Opening up the proscenium

A posed shot of the multiple-camera teams for Sunny (1930).

In his autobiography Capra regarded The Donovan Affair as a turning point in his career. While complaining about the constraints on sound cameras, he regarded the film as “the beginning of a true understanding of the skills of my craft: how to make the mechanics—lighting, microphone, camera—serve and be subject to the actors.” To my way of thinking, what he did within the confines of early talkies was inject some of the pictorial fluidity and impact on display in the silent cinema. The Donovan Affair is a lot less pictorially stilted than most of the 1929 American films I’ve seen.

Like many early talkies, the film relies heavily on multiple-camera shooting of the sort still used in TV comedies and soap operas. For this film, according to Joseph McBride, five cameras were used. The range of coverage allowed for continuous sound recording while preserving the option of analytical editing. While one camera recorded a wide framing, a long lens could scoop out a closer view without moving that camera into the space.

So when Mary calls on Donovan in the second scene, we can see her arrive in a long-lens mid-shot, then we can see Donovan’s reaction in a long shot. Mary comes forward and a third camera picks them up in a two-shot.

          

The technique permits smooth matches on action, as when Mary turns her head in a medium shot and leaves in the background of the master shot.

     

But it’s also clear that multiple-camera shooting allowed for alternative takes. The mismatches we can detect across some cuts suggest this. During continuous dialogue, Lydia’s arm is up in one shot and down in the next.

     

Capra claims that Columbia allowed him to print only one take, but he got around this by not slating his shots. He simply let the camera run and replayed the action, sometimes several times. This allowed actors to improve their performances in a fluid process. CApra wound up with only one take number, out of which he could pick the take he wanted. This tactic might explain disparities of figure position.

Multiple-camera shooting could yield some awkwardness, especially when shot scale wasn’t adjusted. Here’s an interesting example showing Jean’s quarrel with her stepmother Lydia. As Lydia crosses in front of Jean, the match cut on my second frame below creates a little bump by not sufficiently changing the angle and figure size. The slight pan following the women accentuates the disparity.

               

A jerky cut like this would be very rare later in the 1930s, as it is today. As recording technology improved, Hollywood would move back to single-camera shooting for most scenes, and shot scales were more exactly planned and executed.

If multiple-camera shooting feels a bit theatrical, it’s because it surrenders a degree of freedom in camera placement. We always seem to be watching things from outside a proscenium, even when items are enlarged. I think that the early talkies’ habit of starting with a flamboyant camera movement, as in Sunny Side Up (1929) and The Broadway Melody (1929), was a way of saying, “Don’t worry, this is still cinema.” Capra follows this habit by beginning on a close view of a pile of Donovan’s IOUs and then tracking back very far to set his first scene among the gamblers.

          

The film’s second “act” starts with a comparable flourish. A long shot of the library shows Carney bustling toward the offscreen front door in the background. The camera tracks in and waits for him to return.

          

Presumably we would be hearing him greeting Inspector Killian offscreen, but the frame is empty for about fifteen seconds. Then Carney and Killian stride in, and the camera pulls back to something like the initial master framing.

          

Another way to add fluidity was to insert shots that don’t require lip synchronization. Cutaways to objects or reaction shots could be inserted into the dialogue-laden stretches. For example, Nelson and Mary peer into the library to watch the guests, and Ted Tetzlaff’s cameras give us two shots from positions different than we’ve seen so far.

     

More boldly, after Donovan is killed, Capra gives us not only a shot of Mrs. Lindsey wailing but brief close-ups of Jean’s and Lydia’s reactions.

                         

These two shots are only 32 and 36 frames respectively, harking back to the punchy montage of silent cinema, and they stand out by contrast with the extreme long-shots around them.

Similarly, you have to think of all those subjective POV shots in 1920s silent films when we watch Killian survey the suspects. As he talks with Rankin, he scans them right to left in a series of POV shots, including one longish, zigzag pan. Here are some extracts.

               

The end of the pan shot shows Lydia shifting her gaze from Killian to her husband, off left. We then cut to her husband and Killian, who’s just finishing his scanning of the suspects. Her gesture cast suspicion on her; after all, she’s hiding her affair with the dead man.

     

Most ambitious of all is Capra’s handling of the dinner table situation. Smoothly integrating singles and two shots of characters around a table was one of the triumphs of classical Hollywood style in the late 1910s. Matching character positions and eyelines from shot to shot became part of every director’s craft. Capra adapts talking pictures to these constraints in a flow of shots that include dialogue and keep all characters’ relationships clear and consistent while picking out important details and reactions.

For example, when Donovan talks with Mrs. Lindsey, we get her husband’s glare as he watches from across the table.

     

Here the proscenium is broken. Instead of covering the action from outside, now the camera penetrates space and can take up a variety of angles among the characters.

This camera ubiquity can be exploited to dramatize the murder weapon. A master shot shows the dinner party, with Rankin at the head of the table and his wife Lydia at the foot, turned from us.

Mary brings the roast and the carving knife to the table. Cut in to Mary lifting the knife, which glints.

     

Donovan flinches as the reflection dazzles him. Cut to Lydia, from the foot of the table, saying, “Mary,” and looking down the table to where Mary would be.

     

We’re in a sort of triangular space—Mary-Donovan-Lydia—and the relationships are reaffirmed when a return to the earlier setup shows Mary responding to Lydia’s look.

The dialogue continues within these shots. Capra’s changing setups must have required quite a bit of effort in 1929, what with cameras in booths and the pressures to stick with continuous multi-camera takes. But, reviving silent-film table coverage, he gives us what would be one conventional way to handle such an action in the 1930s.

There are other felicities in the film, but I think I’ve said enough to indicate how it’s an enlightening transitional work. Tied to the multi-camera technique for most of its running time, Capra breaks away for significant stretches. He found a way to give camerawork and cutting some moments of the sort of fluidity that would pervade Hollywood in the 1930s.

 

Capra wasn’t the only major director to cut his talkie teeth on adaptations of stage thrillers. Hawks made The Criminal Code (1931) from a 1929 prison drama, while Hitchcock based Blackmail (1929) and Number Seventeen (1932) on popular plays from 1928 and 1925. Film mysteries would change in the 1930s, with sophisticated detective comedies like The Thin Man and adaptations of the adventures of Charlie Chan and Perry Mason.

Creaky though it might seem in later decades, the manor-house murder would remain a reference point. On stage its premises were given a new twist in Christie’s The Mousetrap (1952), enhanced with social criticism in J.B. Priestley’s An Inspector Calls (1945), and parodied in Tom Stoppard’s The Real Inspector Hound (1962). You can argue that the provincial procedurals with a British accent (Wexford, Midsomer Murders, Inspector Morse, and so on) grow out of the spillikins-in-the-parlor tradition, but shifting the viewpoint to that of the detectives.

---

Thanks to Lynanne Schweighofer, Zoran Sinobad, and Mike Mashon of the Library of Congress and Jim Healy, Mike King, and Roch Gersbach of our Cinematheque for arranging my viewing of The Donovan Affair. Thanks also to Ben Brewster for discussions of the film and 1920s theatrical practice.

A fleshed-out synopsis of the film was published at the time.

Bruce Goldstein had the excellent idea of adding live performers to screenings of The Donovan Affair. He details his efforts to find the original dialogue in this TCM article.

Ray Collins fans may be interested to know that in the stage version he played the butler–described in the script as “really the only acting part in the play.”

My quotation about rapt audiences for the stage version comes from the review, “’The Donovan Affair’ Thrills in Mystery,” The New York Times (31 August 1926), 15. Quotations from Capra are from the first edition of his autobiography, The Name above the Title (New York: Macmillan, 1971), 105. Joseph McBride has written the most comprehensive and probing biography, Frank Capra: The Catastrophe of Success (New York: Simon and Schuster, 1992).

In his autobiography Owen Davis remarks that the failure of The Haunted House, a “burlesque mystery melodrama” impelled him to write The Donovan Affair. That was a play that “could hardly be accused of not being mysterious enough, and one that actually dripped with blood and missed no trick at all of the tried and true tricks of mystery story writing from Gaboriau through Poe and Doyle and Anna Katherine Green and Mrs. Rinehart, even taking in the tricks of the present, and very skillful, crop of hard-boiled detective story writers who were at that time printing their first compositions on a slate in some primary school.” See My First Fifty Years in the Theatre (Boston: Baker, 1950), 94-95.

My exploration of theatrical thrillers has been aided by Amnon Kabatchnik’s two excellent reference books: Blood on the Stage: Milestone Plays of Crime, Mystery, and Detection: An Annotated Repertoire, 1900-1925 (Scarecrow, 2008) and Blood on the Stage, 1925-1950: Milestone Plays of Crime, Mystery, and Detection: An Annotated Repertoire (Scarecrow, 2010).

As ever in such matters, Mike Grost’s encyclopedic website offers background information and critical discussion of authors, periods, and conventions.

On multiple-camera shooting in early sound film, see Chapter 23 of The Classical Hollywood Cinema: Film Style and Mode of Production to 1960. A good primary source is Karl Struss’s article “Photographing with Multiple Cameras,” TSMPE XIII, 38 (1929), 477-478.

The Donovan Affair (1929).