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

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

Archive for the 'Narrative strategies' Category

Nero, Archie, and me: Preface to a new web essay

“Frame-Up for Murder” (“Murder Is No Joke”), Saturday Evening Post (21 June 1958).

DB here:

“It is impossible to say which is the more interesting and admirable of the two.” Thus Jacques Barzun on Nero Wolfe and Archie Goodwin, the characters created by Rex Stout. I’d go farther. For me they are among the outstanding figures of American fiction, and Stout is one of the finest American writers of the twentieth century.

Wolfe and Archie are detectives, and Stout wrote mystery fiction. But that just goes to prove that genre work isn’t just good “of its kind.” It’s often good, period.

Stout started as a pulp novelist. He moved to what we now call literary fiction before pursuing a variety of more mainstream genres. He wrote quasi-experimental novels (one is told backward), comic romances, lost-world sagas, and a political thriller, The President Vanishes. (Stout admitted that Wellman’s film version was better than the book.) He introduced several other detective protagonists, but he wisely concentrated on Wolfe and Archie, whom he launched in 1934 and traced through dozens of adventures until his death in 1975.

Here is Stout with his wife Pola and daughter Barbara in 1935.

Are these stories still read? I hope so. When I quizzed the grad students in my seminar this spring, most seemed unaware of what for me–and, I think, other baby boomers–were major figures in popular culture. It’s partly because the ancillary media never did a good job spreading the word. The film Meet Nero Wolfe (1936) was weak, in keeping with the general failure of 1930s films to capture the strengths of supersleuths like Philo Vance and Ellery Queen. (Better were the Mr. Moto and Charlie Chan vehicles.) During the late 1990s, Bantam reprinted the books in quality paperback editions, with introductions by other writers (Westlake, Lippman et al.). There was a solid cable-TV series (A Nero Wolfe Mystery) back in the early 2000s, with a trying-a-bit-too-hard Timothy Hutton as Archie but a near-definitive Wolfe in the form of Maury Chaykin. But that’s all pretty far back now.

O well .  . . . If Perry Mason can come back, maybe Wolfe and Archie can too. And they won’t be scruffy and unshaven. Archie explains: “I was born neat.” So was his prose.

Today, I’m posting a long (22,000 words!) essay on Wolfe and Archie in an effort to show some reasons they and their creator matter. Background follows below.

 

Research? Say rather, obsession

Like most academics, I try to pour my obsessions into my research. My interest in mystery stories (fiction, film, even theatre and TV) is surfacing in my current project, a book studying principles of popular narrative as they emerge in thrillers and detective stories. In a way, the book turns inside out what I tried to do in Reinventing Hollywood. There I traced film techniques to sources in adjacent media. Now I’m looking directly at those media to consider their formal and stylistic strategies, with some consideration of how film shaped and was shaped by them.

My survey covers familiar ground. I consider the classic detective story (the so-called “Golden Age” of Christie, Sayers, Queen, Carr et al.), the hardboiled writers, the suspense thriller, and the crystallization of major techniques in the 1940s (a polemical point with me). Later chapters analyze the heist plot, the police procedural, and particular writers (e.g., Cornell Woolrich, Ed McBain, Richard Stark, Patricia Highsmith). As presently planned, the book ends with an odd-couple pair of chapters: the “puzzle film” trend of the 1990s, and the domestic thriller of recent years (Gone Girl, The Girl on the Train). Some chips from the workbench have littered this blog, as the links indicate.

One of my key questions is: What has enabled readers and viewers to welcome the sort of extreme complexity that we find in movies like Pulp Fiction or Memento? We’re all now connoisseurs of narrative intricacy, in not only crime plots but science-fiction and fantasy ones (which actually often rest on mysery premises; see Devs).

Another theme of the book is one I pitched in Reinventing Hollywood. I argued that very quickly “modernist” innovations in fictional technique were channeled into something more moderate and middlebrow. It’s there that wider audiences gained proficiency in handling flashbacks, ellipses, plunges into subjectivity, opaque exposition, and other challenges to linear storytelling. At the end of the 1920s, while Joyce and Woolf and Faulkner were pushing modernism into ever more recondite byways (Finnegans Wake, The Waves, The Sound and the Fury), other writers were adapting advanced techniques to traditional ends.

And traditional genres. I’ll argue that modernist experiments provided some strategies that mystery writers could recast within the bounds of their tradition. Stout’s first two serious novels, How Like a God (1929) and Seed on the Wind (1930) joined the trend toward mainstreaming modernism. By 1930 he had fully absorbed what the 1910s-1920s writers had achieved. What’s less obvious is the mark of modernist writing on his reader-friendly detective stories. Yet one lesson of modernism, I think, is the need to play with language, to “defamiliarize” literary texture. That, I think, is one of the tasks Stout took on in the Wolfe saga.

I had planned to include a chapter on the Wolfe/Archie series, but what resulted was far too long in comparison with its mates. So I’m posting it online as an essay. A shorter version will show up, somehow, in the finished book.

Luckily, thanks to the Internetz, the essay’s web incarnation can include some illustrations from magazine publication of the stories. Since most of the books’ drama takes place in Wolfe’s office, artists are challenged to find nifty ways to show people grouped around a desk.

I hope, then, that fans of the series will find some ideas and information of interest. Above all, I hope what I’ve said here and there will stimulate novices to venture on one of the great exercises in American popular storytelling. If you don’t care for it, all I have to say is: Pfui.


There are many good websites devoted to mystery fiction. The most comprehensive is Mike Grost’s. You should also look into Martin Edwards’ fine blog Do You Write Under Your Own Name? See also the official Wolfe Pack site, full of links and assorted Stoutiana. John McAleer’s biography Rex Stout (Little, Brown, 1977) is vastly informative. Stout was not only a great writer but also a fascinating man.


Fer-de-Lance, serialized as “Point of Death” in American Magazine (November 1934).

Who will match the Watchmen?

Watchmen (2019), Episode 7.

DB here:

If we want to understand the craft of filmmaking, we should pay attention to what filmmakers say about their technique. Sometimes we have to confirm what they tell us with other evidence and the onscreen results, but practitioners usually offer unique angles on their creative choices. For this reason I found it fascinating to come upon what Director of Photography Gregory Middleton had to say about his work on HBO’s limited-series Watchmen.

I had read the comic back in the 1980s and was intrigued by its experiments in pictorial storytelling. I’d also seen the movie, which I thought not as adventurous as the book and vitiated by the book’s adolescent nihilism. (Your mileage may vary.) But Middleton’s interview made me watch the series. So I thought I’d use this anxious pause in normal activities to share some ideas with you.

Middleton has a lot to say about the visual design, such as the sort of lenses he used to get startling depth effects. These echo some common graphic-novel compositions.

     

But I got more interested in other questions. How does the show adapt some classical cutting techniques? And how do those techniques square with the “editing” techniques we get in the original comic?

Here’s the relevant passage:

And transitions are the biggest tell when you cut from a beat of one thing to a beat of something else. In the case of the comics, one of the transitional devices we wanted to use to echo the graphic novel is the match cut. In the comic you’ve got one panel with a character in the foreground and someone in the background. The next panel is the same character in the foreground, but they are somewhere else — or somewhere else in a different outfit. And the background is something similar, but it’s something else, and you’re basically just jumping time because you’re really with the character and where their state of mind is, so the intervening time between how they went from here to there is irrelevant.

There is a great one in episode two where Angela walks back after they just sat down, and she sort of disagrees with the idea of going to rustle up the people at Nixonville, and she walks away, and she’s pacing as a match cut to her pacing at the police lineup at Nixonville. You go right where she’s thinking, like Okay, I don’t want to be here, but now I’m gonna be here, and suddenly you’re right there.

The match cut is also a way to instantly transport the audience without introducing a lot of other extraneous visual information. In episode four, we did her opening and closing her trunk when she is going to just get rid of the wheelchair she’s cut apart. She chucks it in the trunk, slams the trunk, and it’s a match cut to her opening the trunk, and she’s already somewhere else. And that just works — you don’t need to see her driving and parking and all that. The point is you just know she is dumping this thing, and it’s a nice, clever way to keep you on point with where she is with her intent.

Afterward, it’s like he’s just jumped ahead again — both places, same time. So it’s a way to sort of double-use that technique. It just takes a lot of planning and preparation. Nicole [Kassell, director/producer] and I worked hard on all the transitions in that episode to try and achieve that effect and make it interesting.

The transitions between scenes that Middleton is discussing are, I think, what have been since the 1940s called “hooks.” His comments open up some ideas about how editing works with them.

There are some spoilers, but not my usual quota.

 

Matchmaking

We can enhance our understanding of filmmakers’ creative options if we systematize a little bit the choices they face and the terms they use. That’s what I try to do in the online essay on hooks between scenes. There I suggest that we can have four sorts of palpable hooks from shot to shot: sound/sound, sound/image, image/sound, image/image. My examples come from National Treasure (2004), a movie that revels in tricky transitions.

For example, one transition in Watchmen creates a sound/ image hook. The Game Warden tells Adrian, “No Mercy” and we cut to a close-up of a Mercy perfume being tested on a focus group.

     

Middleton calls the cuts that create image/ image hooks “matches.” “Match” is a broad term that points up a general principle of cutting for continuity. In a match shot-change, which might be a cut or a dissolve, something in shot B picks up something that was in shot A. Without distinguishing different types of matching, Middleton is stressing how continuity editing techniques, normally applied within scenes, can be used as transitions between scenes as well.

So let’s create an initial menu of possibilities. Within scenes, there’s the eyeline match, which shows a single character looking, followed by a shot of what the character sees (not necessarily from their optical viewpoint).

Shot/reverse exchanges like this one often rely on the eyeline match, but need not; you can have a shot/reverse pattern if the characters aren’t looking at each other. (Imagine two people talking with their backs to each other.) The eyeline match can link more than two characters within story space.

     

     

     

There’s also the match on action, where a bit of movement in A is carried over into B, despite the change of camera position. The scene between Wade and Angela continues with a combination of the eyeline match and the match on action.

     

Another sort of match plays upon the sheer pictorial qualities of the two shots. As Middleton says, shapes and color patches can be carried over from the background or foreground or both. Filmmakers don’t seem to have developed a specific term for this type of shot change, although they were clearly using it from at least the 1920s onward. (Eisenstein was the most explicit.) In Film Art: An Introduction, Kristin and I proposed calling the lining-up of compositional features from A to B a graphic match.

Precise graphic matches are common in experimental films, but they seldom show up within scenes of narrative films. Still, Eisenstein, Lang, Ozu (here, from An Autumn Afternoon), and a few other directors made bold use of them. And they’re obviously a matter of degree: Ozu can match an overall shot in some detail, as above, or just one portion of the frame.

We might add the possibility of the “category match,” as when a prop in shot A is followed by something comparable in shot B. Imagine cuts among wanted posters inside a sheriff’s office.

When we move beyond the individual scene, what are the possibilities? Clearly, all types of match-cutting can be used for transitions.

In Watchmen, an eyeline match shifts us from Angela, lying wounded on her living room floor, to a view of Judd looking down at her in the hospital when she awakens. Here the eyeline yields an optical POV.

     

A match on action takes us from one scene of Cal opening his hand to another as Angela reaches for the hydrogen-atom ring.

     

A hook can also exploit a graphic match. As Laurie lies sideways after a bomb blast, the image dissolves to a similar composition, a twisting view of a masked bust.

          

Middleton points to one striking moment that viewers are likely to notice: a match on action that’s also a bold graphic match. Angela strides away from one scene and continues her movement, and her compositional thrust, in a new location, Nixonville.

     

The “category match” is at work on occasion too, as when one scene ends on a pocket watch and the next starts on a clock face.

     

The category match blends with the graphic match when hands links scenes.

     

What seems to have sensitized the show’s creators to the possibilities of hooks is the fact that the Moore/Gibbons comic makes punchy use of graphic matches between panels. These are often used to signal character memories by opening and closing flashbacks.

Vivid as these examples and others are, the hook, even the graphically matched one, has a long lineage, as my essay tries to show. The creators of Watchmen are working in a distinct tradition of classic filmmaking.

 

Showing and showing off

More broadly, that classic tradition favors speed and quick pickup. The new Watchmen is essentially a police procedural, unlike the comic and the feature film. Accordingly,  the investigation of a crime is interwoven with the personal lives of the police, the “cop soap opera” that brings in friends, family, lovers, and secrets from the investigators’ pasts.

The crucial backstory involves Angela Abar, who knits together many threads. The child of a US black soldier and a Vietnamese woman, she became a policewoman in Tulsa and underwent the new militarization of the force in response to white supremacy. Some of the earlier Watchmen team, notably Adrian Veidt and Dr. Manhattan, are still exercising their power, while Laurie Blake, the daughter of Silk Spectre, is now working for the FBI.

Like the comic, the series must move the plot swiftly between past and present. The comic splinters time more audaciously, jumping back and forth across its grid of nine vertical panels. But in a classical film, where you don’t normally pause or go back, the story must unwind linearly, and so other strategies come forward to impose a felt pattern.

Older films might dwell on the aftermath of a scene, letting the tension relax a little, and provide a cadence that marks off a distinct section. Modern films, especially if they’re presented in the distracting environment of home viewing, display a need to maintain a rapid pace. So scenes push to a high point, then provide a “button” like that close-up of Laurie, and then move swiftly to the next. Among other advantages, this tactic keeps people from pausing or changing the input. Still, the scene shifts need to be clear fairly quickly.

Hooks can help on both fronts. They snap up the pace, keep us alert, maintain interest from moment to moment. A less clear-cut hook can provide a bump of surprise that teases us into the next scene. Soon, though, we will get imagery, and especially dialogue, that anchors us in the ongoing action.

Hooks also provide motifs that can cascade through the film. Adrian’s ambitions to be the new Alexander the Great are provided by cuts that mock his pretentions.

     

And, as in the comic, the cinematic hooks facilitate subjectivity. The most common way classical storytelling justifies unrealistic or nonlinear narration is through getting into characters’ heads. (Middleton: “You’re really with the character and where their state of mind is.”) Abrupt and startling cuts can be motivated if what we see is a memory, a dream, or a hallucination. This is what we get when Angela embraces Judd’s legs: a match-on-action hook to a flashback of her slow-dancing with Cal.

     

The most elaborate hooks can provide something more: a certain elation in virtuosity. The precise matching of Angela’s stride across those two scenes above is a stylistic flourish that asks us to appreciate the care and bravado of the filmmakers.

Watchmen‘s hooks, however flamboyant, stand out against a pervasive backdrop of classically constructed scenes. We get lots of over-the-shoulder reverse angles, properly timed reaction shots, now-standard cuts to extreme distant shots that provide a “beat” for the scene, and other devices that have been around for decades. Plenty of other traditional strategies for narrative coherence are on display, from goal-oriented protagonists and deadlines to newspaper headlines and TV broadcasts as sources of impersonal narration. There’s even the inevitable Hollywood line: “What are you doing here?” As usual, striking moments emerge out of a familiar flow of narrative cause and effect, with some mysterious gaps eventually filled.

 

Paging the Watchmen

Comics artists absorbed the lessons of continuity filmmaking. If the hooks in films stand out by contrast with the overall smoothness and linearity of classic construction, something similar happens in comics. Here are three “continuity shots” from the Watchmen graphic novel, complete with obedience to the axis of action, reverse-angle cutting, and a “single” for emphasis on Rorschach’s reaction.

Even a match-on-action hook can be suggested, as when Laurie’s lifting a goblet on Mars “cuts” to her as a teenager lifting a dumbbell.

Graphic continuities and discontinuities can be motivated by science-fiction premises, as when Dr. Manhattan teleports Rorschach out of  the facility. The sense of instantaneous change wouldn’t be so strong if Rorschach’s size and placement in the format weren’t kept constant.

In a way, the graphic continuities are almost needed for the time-shifting transitions. A new angle on Adrian at the start of the funeral flashback below would be more confusing than the frontal close-up that completes a gradual enlargement of his face. The new costume and the “offscreen” dialogue cue us to a new scene.

To keep things tidy, the flashback ends with a foreground graphic match on Adrian, leading us back to the funeral.

As ever, popular narrative maintains symmetry and redundancy to keep the action clear. Here’s a more “staggered” but no less symmetrical foreground graphic match–used, again, for subjectivity.

The rectilinear staging evokes either a track-in (the first two panels) and a track-back (the last two) or axial cuts in to and back from Dr. Manhattan. (The Soviets called these in-and-out edits “accordion” cuts.)

So there are some rough equivalents between film editing and the principles binding comic panels, especially ones this jammed with information. But comics have unique resources too–ones that Watchmen‘s creators were well aware of. Dave Gibbons notes that he exploited techniques articulated by the great Will Eisner, “where you could, say, hold the camera’s point of view and have characters walk past it, or follow the characters through a scene acrosss a changing background, where you could echo panel compositions . . . .”

The echoes Gibbons mentions depend on the fact that the book page and its role in a two-page spread create a force-field that displays the story both as a timeline and a spatial array. Panels are at once phases of action and a simultaneous order of images, and the interplay of story information and pictorial pattern can become quite rich, even baroque.

Consider when Rorschach rummages in the Comedian’s closet. Even if we could replicate the panels in a movie shot by shot, the page’s severe structure couldn’t be grasped as a simultaneous whole onscreen.

The echoes are striking. Panel 1, the establishing shot, is enlarged in panel 7 on the bottom row, as is panel 2 in panel 8. Each lower-tier mate is markedly closer than the top one, emphasizing the growing sense of discovery. Panel 2, with its sidelong strips of solid black, is echoed in panel 4 and panel 6 (a slight “pan” left from 4) and, with another enlargement, in panel 9, the climax. The color slabs (yellow/black, grey/purple) accentuate the repetitions and variations. The middle stretch, panels 3-6, forms a suite of variations as well (5 picks up 3, 6 picks up 4), while Eisensteinian cuts flip us back and forth across the axis formed by the straightened coat hanger. All the cuts are compass-point shifts, moving us in 90-degree multiples. The crisp geometry of the découpage renders the unhurried professionalism of Rorschach’s search, an effect accentuated by the snapshot imagery (no motion lines).

The effect depends in large part on the presence of all these images on the page at once. The alternation of shot scale and color fields can be discerned only by seeing the overall architecture of the page.

Panels 1, 3, and 5 build to the climax of the wide framing on the bottom as we pull farther back from Dr. Manhattan during his soliloquy. Panels 2 and 6, echoes of a close-up of the photograph, provide a sharp contrast to the landscape and emphasize the power of memory. That power is enacted in another variant of the hand image, as the young Jon Osterman arranges watch parts, like stars, on a velvet ground.

The page is one standard mid-size unit for the comics story, but the two-page spread is the next phase up. A continuous action can be spread across the open book, with echoes emerging. In  the big fight scene, the shifting POV (occasionally objective, but also bouncing us among Rorschach, Adrian, and Daniel) is capped by symmetrical wide panels at the bottom of each page. The rose-tinted panels on the right page add their own fillip of pattern.

     

In a film such symmetries and variations unfold over time. We can make them apparent only through a “spatialized” analysis, spread out on the page like a comic-album layout. (Such were those pioneered by Raymond Bellour in the 1960s and 1970s.) In comics, these patterns strike us immediately, as a primary and palpable expressive vehicle. One-off devices like the graphic match can be sustained both on the screen and on the page, but each medium stamps its own structures on the flow of time.

 

There’s a lot more chop and change in the space-time continuum of the Watchmen comic than I can do justice to here. (Some pages run three time periods simultaneously.) But my main point should be clear. As usual, when we start to pay attention to the films and media artifacts that fascinate us, we always find more going on than we realized. Both movies and comics work on us so directly that their design subtleties can easily pass unnoticed. We can keep reverse-engineering them as long as we like, and we’ll learn a lot. And the creators can be very helpful guides about what to pay attention to.


I first became aware of Gregory Middleton’s comments by reading the review essay by Namwali Serpell, “In the Time of Monsters,” New York Review of Books (9 April 2020). I’m grateful to Serpell for calling my attention to the problem of cutting in the Watchmen series.

The essay is somewhat misleading in its account of match cutting, which is said to create “‘jumps’ rather than flows from shot to shot.” It’s actually just the opposite. The match cut creates continuity of time, space, and graphics between shots. It need not, as the essay claims, link two settings, and it’s not an alternative to a “splice cut,” which is a term with no currency I’m aware of. (A splice is simply the physical join between film strips, and it can govern any type of cut. Splices are increasingly rare with the dominance of digital editing.) The alternative to match cutting would be discontinuity cutting.

Serpell’s essay is worth reading for its interpretive dimension. She argues that for all the creators’ efforts to create strong images of minority heroes fighting white supremacy, there is a critical, antiheroic aspect to the original comic that isn’t picked up in the series. The sanctioned violence of Angela, Wade, and others is treated as unproblematically righteous.

I’d largely agree. Yet I’m not sure that the original book’s portrayal of superhero trauma (the result of the go-to source, childhood abuse) elevates these caped crusaders much. They seem to me best understood as part of that cycle of debunked superheroes we find in other 80s comics, notably The Dark Knight. Why would someone put on a mask to fight crime alone? What dignifies vigilantism? Are these “the heroes we need now”?  These, we’re told, are deep questions, but they seem to me the tropes of a quickly standardized set of conventions. They signal a “new complexity” and “adult attitude” comparable to the romanticism of film noir and the cynicism about samurai ideals we find in Japanese swordplay movies. Bringing heroic genres low is itself a genre convention.

More broadly, I’d propose that the the storytelling strategy Serpell highlights is part of the common move within thematically ambitious mass-market movies to be strategically ambiguous. Creators, I’ve suggested here, often throw many things against the wall to see what will stick. They can then point to contradictory aspects which show balance. So yes, Angela tortures members of the Seventh Kavalry, but (a) they’d do the same to her; and (b) she’s driven by righteous vengeance for ancient wrongs; and (c) we want to make you a little uncomfortable with her vigilantism anyhow. Nowadays, it may be that what seem involuntary contradictions are deliberate efforts to have things many ways at once. This maximizes viewer outreach and gives critics fodder for debate. For another recent instance, see the previous entry on The Hunt.

The Dave Gibbons quotation comes from the supplement “The Phenomenon: The Comic That Changed Comics” available on the DVD of the Director’s Cut of Watchmen, 9:35-10:14.

Raymond Bellour’s “spatialized” analyses of repetition and variation can be found collected in The Analysis of Film (Indiana University Press, 2001). Some efforts of mine are in On the History of Film Style and in Ozu and the Poetics of Cinema (available for free, if patient download here), as well as in other material on this site.

Watchmen (2019).

When Hollywood ruminates: Calm after THE MORTAL STORM

The Mortal Storm (1940).

DB here:

We don’t usually think of classic studio cinema as particularly contemplative. But many films open up spaces for quiet reflection on what we’re seeing, or have seen. In the 1940s, that tendency owed a good deal to the ways that Hollywood became, to put it roughly, more novelistic.

Movies had been based on novels for many decades, but in Reinventing Hollywood: How 1940s Filmmakers Changed Movie Storytelling, I claimed something more specific. I think that many 1940s filmmakers became more acutely aware of sound film’s capacities for manipulating time and point of view. These are novelistic techniques par excellence, distinctly different from the largely “theatrical” conceptions of presentation that we find in most studio cinema of the 1930s. Films turned inward, probing perceptions, memories, and dreams. If you need a storytelling twist, one journalist cracked, just call it psychology, and it will get by.

Of course, there are plenty of precedents for time and POV shifts in early decades. Still, I wanted to show that between 1939 and 1952 many major options emerged and were developed in ambitious directions. The results left a legacy. Today, when flashbacks are common and we easily grasp elaborate shifts in viewpoint, our films rework the possibilities refined and consolidated in the 1940s.

One stretch of the book considered some films from early-to-mid 1940 that offered glimpses of innovations we’d see in later years. Married and in Love, One Crowded Night, Stranger on the Third Floor, and Edison, the Man previewed techniques that would dominate the next decade. Particularly chock-full of storytelling ideas, I argued, was Our Town, a daring transposition of 1930s theatrical devices into the key of cinema.

Missing from my roster, though, is The Mortal Storm, a film that went into release in June of 1940. I had neglected to rewatch it when writing the book. A couple of weeks back I caught up with it during a screening at our Cinematheque. Restored by UCLA, it fairly shimmered off the screen.

How could I not have remembered the stunning closing sequence? It’s a bundle of narrative strategies that would get elaborated in the years to come. And it achieves its effect from being somber, slow, and based on human absence. In this epilogue, the movie pauses to think.

To talk about this passage and its counterparts, I have to parade plenty of spoilers. Sorry. But at least you get video.

 

Emptying the nest

The Mortal Storm.

The plot starts in Germany at the moment of Hitler’s ascension to power in 1933. Elderly Viktor Roth is a much-loved Jewish professor of medicine. His household includes his stepsons Otto and Erich, his wife Emilia, and their daughter Freya. Two other young men are close to the family: Fritz, who aspires to marry Freya, and Martin, a farm boy training to be a veterinarian.

Otto, Erich, and Fritz become enthusiastic Nazis, but Freya draws closer to Martin, who quietly resists the growing bigotry in their small town. Professor Roth is arrested and killed. Martin and Freya, now deeply in love, try to escape by ski to Austria. A squad of soldiers, under Fritz’s direction, fires on them. Freya is killed.

In the crucial scene, Fritz returns to the family home to tell Otto and Erich.

The sequence blends several narrational choices. Most obvious is the camera movement that seems to drift off on its own. It is, we might say, semi-subjective. It suggests Otto’s slow walk through the now-empty home, but it isn’t clearly marked as his optical POV. The opening stretch of the shot shows him pacing more or less obliquely to us, before the camera pans and tracks to the table and beyond.

     

     

The framing insistently keeps Otto offscreen. Only the sound of his footsteps and pauses suggest that he’s present, off left or behind us, at each station of the shot.

As the camera drifts along, we get auditory flashbacks to earlier scenes, three at the table, and one echoing the Professor at his lectern. Flashbacks are characteristic features of 1940s cinema, and purely auditory ones will come into prominence, as in The Fallen Sparrow (1943). Since we’ve seen the action these flashbacks evoke, they’re as much flashbacks for us as for Otto. Indeed, Otto has been a minor character in the story. His walk triggers them, but his absence from the frame makes it easier for us to project our memories onto these spaces.

The transition among the flashbacks prepares for Otto’s defection. The voice shifts from Freya, whose death Otto is grieving, to the men challenging authority. We hear Professor Roth urge youth forward, and Martin advocating peace and free thought. These later moments seem to crystallize Otto’s change of heart. Dwelling a bit on the statuette of Youth carrying the torch of knowledge suggests that Otto may become inspired to take up his father’s commitment to humanism.

At the climax of the shot, the camera frames the staircase (another icon of 1940s cinema) and starts to back up. This can hardly be Otto’s optical viewpoint. The rapid footsteps suggest that he has left the house by striding out, as it were, behind us. The sound of a closing door confirms our inference. He has left us behind.

As he does in the final moments, when Otto’s flight is depicted as footprints in the snow. Perhaps, now that he has become revolted by the cruelty of Fritz and the Nazi regime, he will take up resistance in Martin’s spirit.

Accompanying the shots of the footprints is the voice-over narrator whom we heard at the film’s start. (Such narrators proliferate in 1940s movies.) His speech, from a poem called “The Gate of the Year,” echoes the visuals: a man “at the gate,” the prospect that one may “tread safely into the unknown.” In giving up Nazism, Otto is giving up home.

The camera tilts up to show the empty house.  The snow-encased home is quite different from the bustling view we saw at the film’s start, when the postman delivered gifts for Professor Roth.

     

Hitler has destroyed the family. Still, as the snow buries the footprints, the narrator urges that divine guidance can provide safety for Otto’s escape, and perhaps a decision to fight Nazism.

We don’t normally think of MGM as a hotbed of cinematic innovation in the studio years. But the company had its moments (here and there), and this is one of them. The Mortal Storm‘s play with time (flashbacks, the solemn duration of the house tour) and viewpoint (Otto triggering some bits of remembered dialogue) resembles what we might get in a psychologically slanted novel of the time. We’re given a few minutes to breathe deeply and think about what we’ve seen, and to build up expectations about what Otto may do.

 

The house as memory vessel

The Miniver Story (1950).

One section of Reinventing Hollywood analyzes Enchantment (1948), a film narrated by a house. The house presents itself as a cozy repository of the memories of several generations. More generally, houses are powerful images in 40s films; think of Tara, Manderly, the estate in Dragonwyck, and the Gothic mansions of Jane Eyre, Gaslight, and The Spiral Staircase. Often they’re presided over by ominous portraits, as Steven Jacobs and Lisa Colpaert have shown.

One lesser-known example is House of Strangers (1949), directed by Joseph Mankiewicz. This story of an Italian immigrant who has become the head of a big bank is framed by his son Max returning to their massive home. The bulk of the film is given in flashback, but the flashback is launched by a wandering camera accompanied by “M’apparti tutt’amor” (from the opera Martha) playing on the phonograph. After a trip up the stairs we are taken via dissolve to the patriarch singing it in his bath.

The plot’s central section treats the lyric (“You all seem to love me”) with doubled implication: Gino’s reckless loans endear him to his customers, but his sons resent his power over them. The long flashback ends at Gino’s funeral, passing from the portrait in the past to the phonograph and to his son Max, brooding under the looming picture.

     

     

The huge Minafer mansion is practically a character itself in The Magnificent Ambersons (1942). One of Welles’ late script versions  envisioned a climax in which George, distraught by the shabbiness of their neighborhood, would pass numbly through the house, and our sense of it would be given through his eyes.

145 FULL SHOT of the Amberson Mansion, seen from behind George who is standing in front of camera. He starts walking toward the mansion. CAMERA FOLLOWS, moving faster than he does and soon is so close to him that his body creates a dark screen for a DISSOLVE TO:
146 CAMERA is on the steps of the Amberson Mansion, MOVING up to the door and STOPPING. George’s hands enter the scene, insert a key in the lock, turn it —
147 On the Narrator’s words, “move out” the door opens and CAMERA MOVES thru it into the house.
MOVING SHOT as CAMERA WANDERS SLOWLY about the dismantled house — past the bare reception room; the dining room which contains only a kitchen table and two kitchen chairs, up the stairs, close to the smooth walnut railing of the balustrade. Here CAMERA STOPS for a moment, then PANS down to the heavy doors which mask the dark, empty library. HOLD on this for a short pause, then CAMERA PANS back and CONTINUES, even more slowly, up the stairs to the second floor hall where it MOVES up to the closed door of Isabel’s room. The door swings open and we see Isabel’s room is still as it always has been; nothing has been changed. FADE OUT

This passage is strikingly similar to what we see in The Mortal Storm. True, the cues for George’s optical viewpoint are more explicit than what Borzage gives us for Otto. But Welles is subtler along another dimension. He counts on our remembering earlier scenes without benefit of auditory flashbacks. The camera revisits the reception area we saw during the ball, the dining room where George challenged Eugene Morgan, and the staircase where George and Fanny quarreled, before coming to rest in Isabel’s room, where we saw her waste away and die. We are asked to supply our own flashbacks.

Some of this POV passage might have been filmed, but it wasn’t retained in Welles’ final version, even before RKO’s mangling. In the film as we have it, only Welles’ lead-in and conclusion remain. The narrator supplies a moving-camera montage, as George registers the changes in Amberson Avenue.

The street sequence ends in darkness and tracks back from George kneeling at the bed begging forgiveness. The fact that the shot starts from his darkened head reinforces the subjectivity of the montage.

Welles’s handling is novelistic in the sense of wrapping a character’s flowing impressions inside an omniscient verbal commentary using free indirect discourse (“Tomorrow they were to move out”). George’s moments of rueful meditation are moments for us as well.

Like the narrator that closes The Mortal Storm, this voice is external to the story world. But the same house-haunting effect can be achieved by a narrator who lives in that world. Coupled with the wandering camera, this can turn our view of a scene in the present into a view of the past–or of an eternal future.

The example I have in mind is from The Miniver Story (MGM again, 1950). In Reinventing, I analyze a sequence that creates layers of time: Clem Miniver’s voice-over in our present, an image of he and Kay in the past, and references in the commentary to periods still earlier than what we see. (The clip of this sequence is online here.) At the end of the film, after the Minivers have married off their daughter Judy to Tom, they must face Kay’s impending death from cancer.

The final sequence shifts from the day of the wedding to a kind of timeless realm in which Kay’s spirit lives on. Again, camera movement suggests an invisible presence.

By 1950, we’ve had several permutations. The camera wanders without verbal narration (music alone in House of Strangers). It does so with auditory flashbacks (The Mortal Storm). It does so with a nondiegetic (external) narrator (The Magnificent Ambersons). And it does so with a diegetic (story-world) narrator (The Miniver Story). I try to show in the book that 1940s filmmakers swiftly expanded, even exhausted, menu options involving many storytelling techniques.

Trust Hitchcock to give us yet another variant, and in a tour de force at that.

 

Enough rope

In the eleven shots of Rope (1948), set almost completely in an apartment, the camera’s peregrinations are usually motivated by character movement or simple track-ins and track-outs. At the climax, though, the camera cuts loose. Sort of.

Brandon and Phillip have strangled their friend David and hidden his body in a chest in their living room. They’ve ghoulishly used the chest as a buffet table for their afternoon party. After the party, one guest, their prep school teacher Rupert Cadell, has returned. He suspects that something bad has happened to David. Rupert challenges the pair and sketches out how they might have murdered their friend.

His reconstruction isn’t wholly correct. David wasn’t bludgeoned, and apparently the armchair played no role. What’s fascinating is that the camera supplies a hypothetical, virtual flashback. As in our earlier examples, David becomes an invisible guest, summoned up by the mobile frame.

The camera traces out the action Rupert posits, emphasizing the hall closet (where Rupert earlier discovered David’s hat) and edging eventually toward the chest. At that point Brandon steps in, with his hand tensing around the pistol in his pocket. He stands in front of the drinks table.

     

That’s the climax of the shot. Cut to a shot of Rupert. He seems to intuit that he should avoid mentioning the chest, so he proposes that they might have carried the body out of the building. As the camera traces out that possibility, Brandon steps back into the frame to confront Rupert.

     

     

Here I think Hitchcock made a mistake. At the end of the first shot, Brandon and his pistol are quite close to Rupert, near the drinks table. But in the followup shot, he’s not visible when the camera pans left past the table to enact the scenario Rupert is considering. Brandon would have had to skip backward like the Road Runner to get as far away as he is when he steps back into the frame in the second shot.

Still, the important effect is the representation of Rupert’s thinking. Goaded by Brandon, he ponders how the crime might have been committed, and the camera carves into space to reveal the scheme he conjures up. A sort of flashback? Yes, but an unreliable replay, left largely to our imagination. Semi-subjective? Yes, since when we cut to Rupert he seems to be staring down at the chest. Yet the camera is too free-ranging to be purely Rupert’s optical POV. He’s not moving around, as Otto is in the Mortal Storm sequence.

In this most “theatrical” of movies, Hitchcock manages to give us a verbal-visual flow that is something like a cinematic equivalent of the novelist’s conditional perfect tense. If Brandon and Phillip had killed David, they could have done it this way. The camera enacts a speculative train of thought. Call it psychology.

 

One thing I shouldn’t ever forget: Hollywood in the Forties is a booming rush of visual, auditory, and narrative ideas. In the approximately 5,655 features released in the period I marked off, there are surely many other startling instances of creative craft. I’ll keep looking.


Thanks as ever to our Wisconsin Cinematheque for fine programming under the auspices of Jim Healy, Mike King, and Ben Reiser and excellent projectionist Roch Gersbach. Thanks as well to Joe McBride for sharing material on The Magnificent Ambersons. More on Ambersons can be found here and here.

The Mortal Storm‘s “novelistic” final moments owe nothing to its source, Phillis Bottome’s The Mortal Storm (1938). An earlier version of my ideas about Enchantment are here.

Chapters 4 through 6 of Patrick Keating’s The Dynamic Frame: Camera Movement in Classical Hollywood offer a careful survey of creative choices facing filmmakers of the period, along with explications of their “practical theories” about cinematography. After writing this entry, I learned that Patrick has also made a fine video essay on the Ambersons sequence.

For earlier blogs on related subjects, see the category 1940s Hollywood.

PS 2 March 2020: Patrick Keating reminds me of another 1940 release I should have mentioned: Hitchcock’s Rebecca. When Maxim narrates his confrontation with Rebecca, the camera moves autonomously to “replay” the scene. It anticipates my Rope example, except that Maxim is recounting what really happened, while Rupert is sketching out his (partially inaccurate) reconstruction of David’s death. Still, though, it’s another variant on an emerging pictorial convention. Thanks, Patrick!

PS 2 March 2020, later: Thanks also to John Belton, who writes to remind me not only of the Rebecca scene but a comparable camera movement in Under Capricorn. Crowdsourcing works.

Rope (1948).

Un-Marry me a little: MARRIAGE STORY and LITTLE WOMEN

Little Women (2019).

DB here:

For a while The Blog conducted an annual ritual of analyzing storytelling techniques in year-end releases. I wrote entries from early in 2016, in 2017, and in 2018. Last year I muffed it, largely because of time spent revising our Christopher Nolan book. (Yes, we’re also looking forward to Tenet, especially after that hellah trailer.)

This time I’m trying an alternative. Instead of surveying a range of releases, I’ll focus on two that I think encapsulate some robust variants on familiar narrative strategies. Those strategies include choice of protagonist, linearity versus nonlinearity in time, and manipulation of viewpoint. While I’m concentrating on Marriage Story and Little Women, I’ll draw out some comparisons with other films.

Many spoilers follow, but of course you’ve probably seen all the new films. Except maybe Cats.

 

Protagonists, dual and dueling

Human nature is not given to a protagonist/antagonist three-act structure. Human nature is just one damn thing after another in which the only thing that matters is what went on today because yesterday is gone. And that is contrary to a lot of the business that we’re in, which makes sure that everybody understands the story by page 30 and is involved in the conflict.

Tom Hanks

You’re plotting a film. What sort of options do you face? A basic choice involves protagonists.

You might build the film around one character who pursues a cluster of goals. Examples this season would include Dark Waters, Motherless Brooklyn, Uncut Gems, and Harriet. The protagonist can have helpers, and will certainly have adversaries, but her or his initiatives, decisions, and responses propel the action. In addition, we’re usually attached to the protagonist’s point of view, which limits us to what she or he knows. Judicious widening of the horizon often takes place to enhance tension. In Uncut Gems we’re briefly attached to Arno’s thugs when they’re tailing Howard, and the climax crosscuts Howard in his office with Julia placing his big bet.

You could center the action on two characters, giving us a dual-protagonist plot. Here the goals may be shared or at least compatible. In The Aeronauts, a lady balloonist and a male meteorologist cooperate, with frictions, to break ascension records, while Ford v. Ferrari unites two men working together to win at Le Mans.

More rarely, a dual-protagonist plot can shift the protagonist in the course of the action. Good examples are Red River (1948) and The Killers (1946). I’d argue that Don Corleone functions as protagonist in the early sections of The Godfather (1972), while Michael takes up that role later. Similarly, Waves initially concentrates on Tyler, but he largely drops out of the plot and his sister Emily drives the film’s second half.

Alternatively, the plot can present two protagonists in competition. This season we’ve had The Current War, centering on the struggle between Edison and Morgan to transmit electrical power. The narrational weight is largely with Edison, but I think Morgan is characterized enough and we’re attached to his viewpoint frequently enough to present a counterweight. Morgan isn’t simply an antagonist but rather what Kristin calls a parallel protagonist, like Salieri in Amadeus or Captain Ramius in The Hunt for Red October. As these examples indicate, parallel protagonists, although they’re trying to figure out each one’s aims and stratagems, often become fascinated with each other and recognize their affinities.

Paired protagonists are common in romantic comedies, which often consist of friction between the couple (due to clashing goals) but end in harmony and union. What’s striking about Marriage Story is that here the end, not the start, of a romantic alliance is treated through the dual-protagonist strategy. Charlie and Nicole struggle over the terms of their divorce, particularly the handling of custody of their son Henry. Unlike Kramer vs. Kramer (1979), which is organized chiefly around the husband’s viewpoint, this gives weight to both spouses.

Director Noah Baumbach achieves this balance through a cunning parallel block construction. The film opens with two montages of roughly equal running time. One surveys Nicole’s habits and accomplishments with Charlie’s voice-over praising her. (“She’s my favorite actress.”) Then we get a montage illustrating what Nicole loves about Charlie, with other incidents stitched together by her voice-over. Both montages weave in scenes of Nicole in rehearsal while Charlie, the director, makes suggestions.

Baumbach has compared these montages to an overture in musical theatre. The film’s score sets out themes associated with each protagonist, and the quirks and routines that rush by establish important motifs, like haircutting, Monopoly games, and Henry’s urge to sleep with his parents. With the he said/she said duality, the montages prepare us for the film’s strategy of parallelism, a compare-and-contrast attitude.

     

The montages are revealed as visualizations of two memoirs the couple have written for a mediator.

     

They’re planning to divorce, and he’s asked them to recall what they loved about each other. Charlie is willing to share his notes with Nicole, but Nicole won’t show hers. This hints that he’s more reluctant to separate than she is, planting a question about why she seems determined to pursue the divorce.

Just as important, we’ve been given privileged access to both characters’ minds, and this sort of alternating omniscience will proceed throughout the film. There won’t be any more plunges this deep into subjectivity, but we’ll always know more than either does, because after they separate we’ll be attached to one or the other in large stretches.

For a time, though, we’re with both. In the mediator’s office, and then during the play’s performance, in the bar with the troupe after the show, and in the family apartment, they interact as a couple. (True, Charlie sleeps on the sofa.) But once Nicole moves to California, the first block of action ends and we are attached to her and Henry as she launches her new project, a TV pilot.

Not until Charlie comes to visit Nicole, her mother, and her sister does the narration bring him back. There he’s officially served the divorce papers. This scene launches a discreet viewpoint pivot from her to him. The family cuddle ends when Henry banishes Charlie from bed, foreshadowing how marginal his father will be to him from now on.

     

     

The film’s next block attaches us to Charlie as he seeks out a lawyer, takes Henry on outings, and clashes with Nicole about how to celebrate Halloween. The couple wind up giving Henry two trick-or-treating trips, in different costumes, which reiterates the duplex structure of action we’ve been presented with since the start.

The alternation between Nicole and Charlie’s viewpoints quickens as their negotiations get more fraught. Their first legal meeting ends with Charlie’s losing faith in his easygoing attorney. The next meeting is an escalating confrontation between Charlie’s new hard-charging lawyer and Nicole’s equally tough Nora. In the courtroom exchange, the he said/she said pattern becomes vicious as each lawyer weaponizes minor incidents from scenes we’ve seen to cast shame on the opposing side.

The nastiness of the custody battle comes to a crisis in a ten-minute duologue in Charlie’s apartment, an all-out fight between Nicole and Charlie. They run through a repertoire of reactions, from assurance of mutual admiration to declarations of annoyance, unhappiness, frustration, and complaints. By the end they’re screaming insults. Charlie rages and then, as if aware of how monstrous he’s being, collapses sobbing at Nicole’s feet.

     

Most classically constructed films follow the pattern Kristin identified back when. The plot consists of a setup, a complicating action redefining the setup, a development section consisting largely of delay and backstory, and a climax that resolves the situation. An epilogue asserts a stable, if changed state of affairs. The four main parts are roughly equal in running time, with the climax tending to be a bit shorter and the epilogue being only a few minutes.

Up to a point, Marriage Story conforms to this architecture. The first thirty minutes set up the split in the family before focusing on Nicole’s new life in California. Both Charlie and Nicole had hoped to separate amicably, with no need for lawyers. But thirty minutes in Nicole hires Nora and sets in motion a more severe legal battle than the couple had expected. The complicating action is triggered by serving Charlie the divorce papers.

There’s no turning back, and the new situation centers on figuring out how to handle access to Henry. Charlie wants Henry to spend time in New York (“We’re a New York family”) but Nicole wants him with her, and as he was born in California the law inclines to her side. Hence the triple thrust of the Charlie block: visiting lawyers, trying to keep his Broadway production on track, and winning some loyalty from Henry.

The development section consists of characterizing stretches (Nicole indulges in a quick sexual encounter) and delays: the unsatisfactory first lawyer session, a power outage at Nicole’s house, and the courtroom showdown. What happens next, though, seems to me quite original.

 

Between theatre and TV

To determine custody, both Nicole and Charlie must let an evaluator visit to observe each one’s treatment of Henry. In a more ordinary film, this stretch would initiate the climax. The visit from the evaluator would furnish a deadline for determining how custody would be handled. Then the film’s peak could be the vicious, trembling argument between Charlie and Nicole. This would be the explosion that reveals both their love and the impossibility of their staying together.

From this angle, Charlie’s guilt-ridden collapse would be the resolution–his realization of how he stunted Nicole’s life. A courtroom finale settling the terms of custody (a little more for Nicole than Charlie) would fill out the climax and lead to an epilogue, perhaps on the courthouse steps.

Excuse me for rewriting the film. I do it to show that Baumbach’s script does something daring. The big argument comes before the evaluator’s testing. After that brutal clash, we see Nicole rehearsing her answers in Nora’s office. Moreover, the blundering efforts of Charlie to convince the stiff evaluator he’s a good father play out in a lengthy comic scene with some gory sight gags.

     

A certain amount of suspense remains, I think, but the final cascade of gags works against the emotional pitch of the couple’s quarrel. Baumbach has, in effect, risked using an anticlimax to round out the normal climax section of the film. It also serves as a good-natured punishment for Charlie’s self-centeredness.

The same daring informs an unusually lengthy epilogue. It’s built out of the sort of modules we’ve seen already. Nicole and her friends and family celebrate her divorce with a party, while Charlie mopes around Manhattan and morosely salutes his play’s closing with his troupe in a bar. We might stop there, but Baumbach again does something original (though highly motivated). Charlie, now relocated for a teaching gig in LA, comes to pick up Henry and discovers the boy reading the note about Charlie that Nicole had prepared for the mediator session.

     

Not only does it reveal the feelings she had suppressed during the session, but the fact that she kept it shows she still harbors affection for that part of her life. Other films surprise us in the epilogue (Citizen Kane, for instance), but Baumbach’s use of the memoir in the film’s final moments remains a pretty bold, and moving, choice. This stretched and packed epilogue shows Charlie how much Nicole loved him, while also suggesting things that contributed to stifling her. Lines like “He’s very competitive” and “He loves being a dad” have a new impact now that we’ve seen his battle for his son.

In telling this story, Baumbach exploits a larger strategy of what theatre people call continuous exposition. Instead of giving the necessary backstory in a lump at the beginning or middle, major information is sprinkled through the ongoing plot. We’re familiar with this device in films that trigger fragmentary flashbacks, filling in backstory bit by bit. Baumbach goes with a more “theatrical” strategy using dialogue to invoke things that happened before the first scenes..

One of the major instances involves Nicole, who breaks down in an embrace with Nora, sobbing that Charlie slept with his assistant. Coming half an hour into the movie, it explains Nicole’s bitterness in the mediation session, as well as her larger reappraisal of her life with Charlie. At other points we learn of big events, like Nicole’s show taking off and Charlie’s long-term settling in LA, in casual conversation, not in extended scenes.

Crucially, in their climactic quarrel, Charlie justifies his affair by accusing Nicole of withholding sex for a year. We can’t appraise the truth of this, but it at least fills in a motive that more conventional exposition would have put into the setup. Resisting the temptation to supply flashbacks for all these revelations, Baumbach trusts our memory. That way the new data can color our ongoing understanding of the characters. The opening montages were generous but one-sided, chunks of incomplete exposition that suppressed important motives and behavior.

Continuous exposition is associated with Ibsen and playwrights who followed, but the theatrical patron hovering over the film is Stephen Sondheim. Greta Gerwig’s Lady Bird had used Sondheim as a touchstone for ambitious high-school players, but the parallel structure of Marriage Story makes more explicit references, this time to Sondheim’s Company. Nicole’s party features her and her mother and sister performing  “You Could Drive a Person Crazy,” a saucy song about dumping a weak man. Soon in the bar Charlie is singing the yearning “Being Alive.”

     

Maybe a little on the nose (like the movie’s title), these citations seem true to the tastes of these show-biz mavens, while suggesting that at least some of Manhattan clings to Nicole in her exile.

Another parallel reminds us of a perennial Hollywood motif. Nicole began her career in a raunchy teen movie but thanks to Charlie’s stage shows she became a respected performer. Yet to establish her own identity more fully she agrees to shoot a TV pilot. Marriage Story positions itself between theatre (a little pretentious, but nobly struggling) and TV (dumb and superficial, but high-tech and well-financed). Worse, TV literally defaces Nicole.

     

Theatre, TV: what about film?

In this story about show-biz LA, movies and references to them are sparse. (I didn’t spot any.) So maybe we should take this film itself as standing in for righteous cinema, rather than the teenpic trash Nicole was in. Perhaps Marriage Story offers itself as its own example of the subtlety and risk-taking that cinema can embody. Even when streaming on Netflix.

 

Muses in the family

The family saga is one Hollywood genre that doesn’t get enough respect. We tend nowadays to celebrate the tough, not tender side of studio cinema. The cult of noir, the abundance of hard-edged action pictures, and the idolatry that trails Tarantino all tend to make us prefer force to gentleness. When families gather, we expect big trouble, if not outright murder (Knives Out). We decry weepies of any sort, and family sagas are often felt to be soft, schmaltzy, womanish. A male friend tells me that Little Women is “a movie about hugs.”  When NPR devotes a whole show to it, panelists ponder how to convince men to see it. No wonder the family film has migrated to daytime cable TV.

Yet the family saga is one of the nicest things American cinema does. Two of our greatest masterpieces, How Green Was My Valley (1941) and Meet Me in St. Louis (1944), are prime examples. The Forties were rich in such efforts, including Forever and a Day (1943),  Life with Father (1947), I Remember Mama (1948), The Human Comedy (1943), and Since You Went Away (1944). We ought to recognize as well the strength of later entries like The Joy Luck Club (1993), How to Make an American Quilt (1995), and Soul Food (1997), all trying out some of the fresh approaches to storytelling that were emerging in the 1990s.

And the sentiments informing domestic sagas seep into other genres. The Fast and Furious team, we’re told, come to be a family, as do the Avengers. The coming-of-age story, that perennial of indie cinema, inherits the aura of cozy warmth that is central to the family saga. I’d add one of my favorites, We Bought a Zoo (2011), which isn’t really a saga but does radiate a comparable warmth.

Although there are probably earlier examples (I think of Vidor’s 1924 Wine of Youth), it seems likely that the 1933 MGM production of Little Women furnished an important template. The four March sisters, each drawn to a different art form, are a model for the musically gifted sisters in the fine Four Daughters (1938).

And surely the fact that Katrin in I Remember Mama chronicles the family’s daily lives owes a lot to the example of Jo March, aspiring novelist.

The family saga poses at least three creative problems for the filmmaker. Since each family member is likely to confront personal problems (romance, finance, school, job) how do you weave and weight multiple storylines? How do you provide conflict to propel the action? And, since the “saga” comparison suggests development over years or even generations, how do you handle long spans of time cinematically?

Greta Gerwig handles all these problems adroitly in her version of Little Women. I’m going to concentrate on the film, but I’m aware that some of the narrative strategies are taken from Louisa May Alcott’s original novel. But much of what’s ingenious about Gerwig’s adaptation is of her own devising.

Start with storylines. In most such films, the trick is to create a group but then produce a scale of emphasis running from minor figures to the most important one typically, the “first among equals.” In How Green, that is Huw, also our narrator; in Meet Me in St. Louis, it’s Esther. But the doings of other characters shape the family’s destiny and the decisions made by the spotlighted figure. So the activities intertwine.

In Little Women, characters shape one another’s development. Jo, the first among equals, is nonconformist and self-reliant. Yet she needs steering–from Friedrich, the professor who tries to turn her away from sensation fiction, and more importantly from Beth, who in her sickness urges her to write “our story.” They give her the strength to persist and trust her sense of what her writing can be.

On the romance front, Jo’s rejection of Laurie’s proposal of marriage opens the field for sister Amy, who has already supplanted Jo in the role of amanuensis to Aunt March. When Jo, out of loneliness, decides to welcome Laurie’s offer, it’s too late: he’s married to Amy. The tangled alliances of melodrama get tightly bound in the family saga.

The other sisters contribute to the causal weave of the plot, with Meg’s decision to abandon the stage reinforcing Jo’s stubborn attachment to her art. More generally, the fates of the sisters dramatize the tension between creative impulse and the social demands of domesticity. Meg wants a family more than fame. Amy, the indifferent painter, can hope only for a good marriage (the same prospect Aunt March makes explicit to Jo).

In family sagas, the siblings are put in parallel. Huw’s brothers leave the household, but he loyally stays, and Rose, the mature sister, has more trouble attracting men than the vivacious Esther. Here, Jo’s kindred spirit is Beth the pianist, whose playing gives solace to Mr. Laurence in his grief. But illness keeps Beth from fulfilling herself either as artist or grown woman. As for Marmee, we’re allowed to catch a glint of Jo’s defiance behind the older woman’s warmth when she confesses that she’s angry every day.

Jo’s main contrast is with Amy. Amy has done nasty things, but she accepts the burden, laid down by Aunt March, of marrying for money, not love, in order to benefit her loved ones. She does this even though, as she reveals in a key scene, she has always loved Laurie and has always felt herself overshadowed by Jo. (Those revelations are also suggested as what leads Laurie to fall in love with her, and not simply as a substitute for Jo.) Even the various suitors get ranged along comparative dimensions of class, strength of will, and temperament.

The need to provide a dense social milieu also creates parallels–here, in terms of good deeds. The Marches are lower middle-class, living on a parson’s income, but they share their Christmas dinner with a more deprived family. The primary family is constantly compared to the wealthy Laurences, who are generous and good-hearted. Even Aunt March, who married well and embraced hardheaded principles, wills her mansion to Jo. The contrast with the flinty publishing house and the imperious Dashwood is softened when we learn, surprise, that he has a batch of daughters himself.

What about conflict in the family saga? There’s often an external threat–predatory capitalism in How Green Was My Valley, the war in The Human Comedy–but not always a personified antagonist. Often these films have no straightforward villains. Parental error can move the plot, as when in Meet Me in St. Louis Alonzo Smith announces that he’s taking a job in New York. And crises are created by misunderstandings or happenstance, most commonly illness. Somebody almost always gets hurt (here, Meg’s twisted ankle, Amy’s plunge through the ice) or sick (Beth’s scarlet fever).

By and large, the conflicts come through romance and sibling rivalry. In Little Women, Meg loves John only somewhat more than she loves fine clothes, so their impoverished marriage nags at her heart. Amy, enraged at not going to the theatre, burns Jo’s manuscripts. Jo responds with hatred–until Amy falls through the ice and needs rescuing. Amy later considers marrying a rich nonentity, and instead acknowledges her long love for Laurie–who in turn loves Jo. As in most melodramas, we know more than any one character, so we watch as characters’ hopes rise against forces they don’t yet realize.

Even without a clear-cut antagonist, the family members can have goals. The March girls start out as aspiring artists, and stretches of the plot are devoted to them developing their abilities. As their goals change, swerving two of them to marriage and maternity, Jo keeps striving toward what we see her doing in the film’s very first scene: selling her stories. Her burning desire to write is a major through-line, and it encounters obstacles of many sorts, from the harrumphing Dashwood to the destruction of her manuscripts by Amy. And even Jo, as we’ve seen, recasts her goals: in offering solace to the dying Beth, she will write “all about us.”

That means writing about the family as it changes over time. Our third narrative problem, in other words.

 

“If I were a girl in a book, this would all be so easy”

Marriage Story‘s avoidance of flashbacks makes it unusual nowadays. It’s hard to find a movie without at least a few flashbacks. As mysteries, Motherless Brooklyn and Knives Out resort to them to replay scenes with new information. Other films make time shifts basic to their architecture. The Aeronauts uses flashbacks to supply the backstory to an unfolding crisis situation, while Hustlers switches between a contemporary interview and stages in the career of the woman questioned.

The Irishman goes for deeper embedding. The overall frame shows elderly Frank Sheeran in the care facility; the next frame is Frank’s trip with Russell Bufalino and their wives to upstate New York, where Frank will kill Hoffa. That trip in turn flashes back to the central story of Frank’s career with the mob. I try to show in Reinventing Hollywood that this sort of Russian-doll structure comes to be a major option in American film during the 1940s.

In Little Women, the years of change in the March family are given through alternating blocks. We start in the present, with Jo in New York struggling to get published. She’s summoned back to Concord because Beth is ill. Meg is living in poverty with husband John, and Amy is in Paris with Aunt March. For about eleven minutes, crosscutting carries us among the sisters.

This block of exposition is followed by a title, “7 Years Earlier,” that sets up the time oscillations we’ll get for the rest of the film. In chronological order we follow the sisters growing up. Chunks of scenes from the past, shifting viewpoint among several characters, alternate with briefer scenes of the ongoing present showing Jo’s settling back into the family. Sometimes the cuts break the blocks into smaller, interlocking bits, as when we shuttle quickly between Beth’s childhood illness and her fatal one years later.

Why split a linear story into two intercut strands? Flashbacks often create a specific sort of anticipation: Not just What will happen next? but What caused the outcome I already more or less know? In the first ten minutes we learn that Laurie proposed to Jo and she refused him; that Amy, not Jo, became Aunt March’s traveling companion; that Meg married John, the impoverished tutor. We’ll witness the development of all these turning points, and more. We must watch the rise and fall of characters’ hopes, knowing they will be dashed. But we also know, from Jo’s initial visit to the publishing house, that she will gain some success. This time-jumping gives us another level of omniscience, one that lets us savor the details of emotional scenes whose outcomes we roughly know. We’re in the theatre, after all, to enjoy the rapture of pathos.

Instead of tagging each time shift with a date, Gerwig expects us to keep track of the double-entry storylines. She assists us by making story motifs visual hooks between past and present. Silk for a dress, a key, Jo seen writing at a window–these link scenes but also carry dramatic weight in the ongoing action. Other echoes are longer-range. We’re invited to remember contrasting dance scenes (tavern, ballroom, porch), scorched dresses, and piano pieces.

Eventually the past scenes catch up with the present. The fusion comes with the burial of Beth and one more clutch of flashbacks, to Meg’s wedding. I take this as the end of the development section. “Childhood is over,” Jo says. Back in the present, Jo vows to abandon writing.

Now the present-time action dominates the climax. Grieving for Beth, distraught at Meg’s leaving the household, and crushed by the marriage of Laurie and Amy, Jo burns her manuscripts–except for the stories she wrote for Beth. She starts to assemble them and write more.

After a glimpse of Dashwood refusing the manuscript, we see Friedrich come to visit the Marches on his way to California. After he’s left for the station, the family claims that Jo obviously loves him.

At this point, in her most daring creative choice, Gerwig retains the crosscutting technique in a way that seems to continue the present/past alternation. Dashwood’s daughters have urged him to publish Jo’s manuscript. In New York, in a scene that rhymes with the opening passage, Jo negotiates with Dashwood.

     

Their conversation is intercut with views of Jo rushing to the station to catch Friedrich and ask him to stay.

The epilogue consists of more alternations. We see Jo watching Little Women being printed, crosscut with a party at the school Jo has founded in Aunt March’s mansion. There John, Friedrich, and Laurie can be glimpsed as Meg and Amy teach children the arts they had practiced. The celebration ends with a birthday cake presented to Marmee.

But there’s another way to take the final minutes. The alternation isn’t tracking two points in time–Jo’s rush to Friedrich and a later session with Dashwood–but rather a split between fiction and reality.

We’re coaxed to take scene of Jo’s pursuit of Friedrich as representing not her own action but the changes that Dashwood demands in Jo’s novel. “Who does she marry?” he asks, explaining that a book sells only if there’s a marriage. Jo reluctantly agrees. “I suppose marriage has always been an economic proposition, even in fiction.” Cut to Jo racing to the station. After the shots of her embracing Friedrich under an umbrella, we’re back in the office. Dashwood suggests the chapter title, “Under the Umbrella,” and Jo agrees. But in turn she makes demands: “You keep your five hundred dollars and I’ll keep the copyright. . . . I want to own my own book.”

We’ve assumed throughout that Jo’s book is highly autobiographical, but we’ve taken what we see and hear as actual events, the living source of a literary text. Instead, the crosscut climax allows a parallel reality to burst forth,  a road not taken. The epilogue yields another ambivalent passage of crosscutting. The school party might be veridical; certainly the family would likely gather for Marmee’s birthday. But the scene could as well serve as the fictional epilogue in Jo’s book (as it does in the Alcott original).

The true epilogue of Jo’s story would then be the moment when she sees her book printed, action that’s crosscut with the celebration. Jo gets the first copy. In the last shot, pleasure, apprehension, and determination play across her features. And she’s framed in a window, an approximate reverse shot to the image that opened the film (see image at the top of this section).

From the start Gerwig has shrewdly foreshadowed the turn to fiction by presenting the film’s title, after Jo has made her first sale, not as an inscription on the screen but as the physical book itself. That book is signed by L. M. Alcott. The apparently identical volume that comes off the printing press at the end bears the name J. L. March. Gerwig has let Jo appropriate Alcott’s story.

     

By giving us a double-voiced ending, Gerwig does something quite bold. Little Women becomes something of a “what-if” movie, positing two paths for her heroine. Alcott’s Jo had given up a literary career, whereas Gerwig’s Jo finds one by writing her own life and adding an optional ending. We’re free to think that Jo and Friedrich married and the family became a harmonious whole, as in Alcott’s book. A happy ending, we might say, for those who want one. But this film about hugs ends with the heroine hugging not a husband but her novel.

Unmarried in life, Jo can marry in fiction, and Gerwig can have it both ways. Narrative lets you do things like that.

 

I haven’t been able to do justice to the intriguing choices made in other films of the season. I appreciate, for instance, the nonlinearity in Kasi Lemmons’ Harriet, where the interruptions of present-time action offer Harriet’s premonitions of future scenes. This sort of “prophetic” flashforward is rare; usually such passages are presented as omniscient narration, not assigned to characters. But the device does establish Harriet as a sensitive, almost angelic figure, and suggests that her quest to guide slaves to freedom is sustained not just by faith but by holiness.

Still, looking in a little depth at just two major films can make us aware of several choices available to filmmakers at this point in history. As Wölfflin said, “Not everything is possible at all times.” But film researchers can usefully trace the flexible menu of options that filmmakers work with, and film viewers come to master.


Gerwig has kindly made available a version of the screenplay, which I discovered only after I wrote this. (Thanks, Kristin.) The notations for the final sequences are pretty interesting. The most wide-ranging discussion I’ve seen of the ending, with plenty of links, is the conversation between Marissa Martinelli and Heather Schwedel in Slate. Among several perceptive reviews, I’d single out the one in Time by Stephanie Zacharek and Richard Brody’s review in The New Yorker.

I can’t help but think how central crosscutting, that technique pioneered by early filmmakers, is to both of these films. Techniques endure because they open up a lot of expressive possibilities.

Kristin elaborates on arguments for four-part plot structure in Storytelling in the New Hollywood: Understanding Classical Narrative Technique. Extended examples on this site are here and here. I discuss family sagas of the 1940s, along with flashbacks, protagonists, and other narrative techniques in Reinventing Hollywood: How 1940s Filmmakers Changed Movie Storytelling. On 1990s revival of these techniques, see The Way Hollywood Tells It: Story and Style in Modern Movies.

Marriage Story (2019). Is this why the film includes few standing two-shots of Nicole and Charlie?

David Bordwell
top of page

have comments about the state of this website? go here