- Observations on film art - https://www.davidbordwell.net/blog -

REINVENTING HOLLYWOOD in paperback: Welcome to the Variorum

[1]

Happy Death Day (2017).

DB here:

Reinventing Hollywood: How 1940s Filmmakers Changed Movie Storytelling came out about eighteen months ago in hardcover. Amazon and other sellers have been offering it at robust discounts [2]. Now there’s a paperback, priced at $30, though that could also be discounted. I hope all these options put it within the range of readers interested in the period, in Hollywood generally, and in the history of storytelling in commercial cinema.

But of course time doesn’t stand still. Since I turned in the manuscript around Labor Day 2016 I’ve encountered some intriguing things that were more or less relevant to my research questions. (I’ve also found a few errors, most of them corrected in the paperback edition. Meet me in the codicil if you’re curious.) In this blog entry and some followups, I’ll discuss some films, books, and DVD releases that came out after I finished the book. They don’t force me to change my case, I think, but they’re things I wish I could have cited, if only in endnotes.

The first entry in this series is here [3].

Seeing Happy Death Day 2U reminded me of one of the central arguments of Reinventing. But before I get to that, let me talk about English drama of the years 1660-1710. No, really.

 

500 plays and more

[4]In the late 1960s a young scholar named Robert D. Hume became curious about Restoration drama. Going beyond the canon, he started reading minor works, eventually discovering a collection of microform cards that included virtually all English plays between 1500 and 1800. The set cost about 10% of his pre-tax annual salary, and he struggled to read the bad reproductions of 17th century printing. Still, the effort revealed something important. “All the modern criticism was so radically selective that the critics had no grasp at all of what was really being performed in the theatre.”

Hume’s method was simple and drastic. He read all of the preserved plays publicly performed in London between 1660 and 1710. Along with revivals, pageants, translations, and adaptations, there were about five hundred “new” plays, and these he concentrated on. They included famous ones like Marriage à la Mode and The Way of the World, as well as many obscure pieces.

The result was The Development of English Drama in the Late Seventeenth Century [5](Oxford: The Clarendon Press, 1976). Its first part surveys the conventions and formulas found in subgenres of comic and serious drama (sentimental romance, horror tragedy, and many others). In the second part Rob traces the development of these types across the period.

From the standpoint of my research I was fascinated to see that Rob reveals a teeming set of variations on generic schemas. Characters, situations, and plot twists are mixed and matched. A plot based on romance leading to marriage commonly shows the couple outwitting blocking characters. Sometimes, though, the man must win the woman over. Or she must conquer him. Similarly, when extramarital seduction drives the action, a man may pursue a woman but

occasionally an amorous, often older woman is the pursuer (She wou’d if she cou’d; The Amorous Widow) or an ineffectual male, usually a henpecked husband, is a comic pursuer (Sir Oliver Cockwood in She wou’d).

In Planet Hong Kong I called this tendency of mass entertainment a “variorum” one, by analogy with editions that print all the versions of a major text.

Once a genre gains prominence, a host of possibilities opens up. Horror filmmakers are likely to float the possibility of demonic children, if only because the competition has already shown demonic teenagers, rednecks, cars, and pets. Similarly, once the male-cop genre is going strong, someone is likely to explore the possibility of a tough woman cop.

You might object that using the term “variorum” is just a fancy way to talk about the mechanical formulas of mass entertainment. But by putting the emphasis on variety within familiarity, the label points up the need for constant innovation, great or small. Fans of film genres readily recognize this churn, but what I began to realize is that it’s common as well in both folk narratives and more “industrialized” ones, like theatre and popular literature.

Monuments of scholarship like Hume’s Development remind us that the variorum principle is a primary engine of popular entertainment. The urge for novelty puts pressure on artists to try to fill every niche in the ecosystem, sometimes forcing competitors to strain for far-fetched possibilities. The wild treatment of noir conventions in Serenity is a recent example.

 

Beyond genre: Style and narrative

[6]

Cover Girl (1944).

Some time in the 1990s I began to realize that a lot of my research applies the variorum principle to domains outside genre. For instance, Figures Traced in Light and the last chapter of On the History of Film Style look at cinematic staging from this standpoint, showing how basic principles of staging got realized in many ways across film history. I applied the variorum principle to particular narrative techniques in some essays in Poetics of Cinema, considering the options of forking-path plots and network narratives.

Reinventing tries to trace a wide range of storytelling options as they consolidated in the 1940s. Rob Hume could study every preserved play, but I couldn’t do that for films. I managed to watch about 600. (Later entries in this blog series will mention some I missed.) Not surprisingly, I found the variorum principle at work in the narrative techniques on display.

I set out a couple of prototypes for the most common plots: the single-protagonist one (Five Graves to Cairo) and the plot based on a romantic couple (Cover Girl). Beyond those, I considered less common plot options, such as multiple protagonists and network narratives. Then I went on to consider narrational strategies that cut across all types. What strategies were available for mounting flashbacks, or expressing subjective states? In effect, I tried to reconstruct Forties Hollywood’s storytelling menu, largely independent of genre.

Another way to put this is that I was tracking norms. But the variorum principle shows that a norm isn’t just a mandate: Do this. Any normative practice is a cluster of stronger and weaker options.

These options lead to a cascade of further (normative) choices. Shoot a dialogue scene in a two-shot and you’ll need to adjust performances for viewer pickup. Shoot using a lot of close-ups and you’ll need to cut among your actors more frequently to keep everybody “in the scene”–that is, salient for the audience.

Choose a flashback and you’re forced to decide how far back to start it, what to include that’s relevant to the present action, and how to remind the audience of action that preceded the flashback. Of course you may also choose to try to make viewers forget, because you’ve misled them. That’s what happens in Mildred Pierce [7] and Pulp Fiction [8].

Usually we find the variorum principle working among several films. What if creators put the principle to work within a single film?

 

The first day of the rest of your life

[9]

Every now and then filmmakers try out forking-path narratives. These plots, I suggest in this essay [10], trace out alternative futures for their characters. Blind Chance and Run Lola Run are prototypes, though there are plenty of examples earlier and later.

Sometimes the protagonist is dimly aware of the options. The protagonists of Blind Chance and Run Lola run seem to learn from their mistakes in the parallel lines of action. This can yield “multiple draft” narratives in which later story lines show characters struggling to achieve the best revision of circumstances they can.

I’d distinguish plots like these from Groundhog Day, which presents not alternative futures but identical replays of a specific time period. What makes each iteration different is that Phil, realizing he’s living the same day over and over, struggles to behave differently. This pattern has come to be called a time-loop narrative.

The time-loop and forking-path patterns usually provide only changes in the story-world elements of each track. Phil’s changing his routines takes place against a background of recurring situations. Similarly, the protagonist of Source Code [11] gets to try out different ways to stop a train bomber.

What, though, if a looped or forking-path movie tried to survey several alternative genre conventions? That would give us a sense that the variorum idea has been swallowed up within a single film.

This happens, I think, in Happy Death Day. By now it’s not a spoiler to indicate that this slasher movie borrows the Groundhog Day premise and loops a single day in the life of mean girl Tree Gelbman. Each day she’s killed by a stalker in a babyface mask. Each time she dies, she awakes in the bed of Carter Davis, who brought her to his dorm room (without sexytime) to recover from a night of partying. As the days repeat, Tree becomes more desperate to avoid her fate and tries a variety of stratagems. They fail, until they don’t. In the course of them Tree learns to become a nicer person.

What’s interesting to me is that several variorum alternatives of the slasher genre are squeezed into this one film. The stalker kills Tree in a shadowy underpass, in a bedroom during a frat party, in her sorority bedroom, under a friend’s window, on a campus trail, and after chasing her through a hospital, a parking ramp, and a highway. She’s knifed, bludgeoned, hanged, run over, and stabbed with a broken bong.

Of course the shooting-gallery premise of slasher films often generates a string of variations across the film. Boyfriends, girlfriends, figures of authority, and passersby are dispatched by Jason or Freddie Krueger in ever more exotic ways. But in Happy Death Day, the sense of genre replay is heightened by Tree’s being the sole target of the ten variant homicides (one of which is a forced suicide). It’s as if we were watching a performer auditioning for screen tests in which she might be cast as one victim or another. But here the victim is always the Final Girl.

The comedy that haunts many slasher films is enhanced by the preposterous premise that Tree will survive. The deaths become vivid as replays by virtue of their tongue-in-cheek humor, as each slaying tries to outdo the earlier ones and as Tree sarcastically comments on her fate.

With the time-loop convention put into place by Groundhog Day, our interest goes beyond changes in the story world and concentrates more on narrative structure. We watch for scenes we’ve already seen, expecting them to be revised in surprising ways. The handling of the replays foregrounds film technique as well, as when in Happy Death Day Tree’s frantic walk across the quad after a late reset is rendered in distorted imagery reflecting her confusion. We register this as a variant on her earlier stride down the same route–hung-over, but not yet desperate.

[12]      [13]

One virtue of such repetitions for low-budget cinema is that the variant passages can be shot quite economically. You can save time on location by reusing camera setups, with the actors altering their performance, or their costumes and makeup. In the DVD bonus material, director Christopher Landon talks of following this production strategy. Why does low-budget cinema explore odd narrative options? They can come cheap.

 

Variants times 2 or 3

[14]

Happy Death Day 2U (2019).

A film series often self-consciously varies the story world that the continuing characters confront. In the studio era, Charlie Chan went to the circus and the opera, Mr. Moto got involved in a prizefight scheme, and Ma and Pa Kettle visited Waikiki. Today’s superhero franchises rely on fully-furnished, constantly changing story worlds. Back to the Future, though, launched a series that did more than present a story world that shifted from film to film. The trilogy self-consciously reorganized its plot structure and narration, with replays and alternative outcomes enabled by a time-travel premise. We’re expected to appreciate the altered replays as part of the film’s experience.

The sequel Happy Death Day 2U has just been released in that Dead Zone [15] in which low-end American genre cinema flourishes. I have a lot to say about it, but it’s probably too soon. Still, it’s no spoiler to indicate that it offers a set of variants on the givens of the first film. For one thing, what caused the time loop of the original is now explained. The birthday motif gets elaborated via Tree’s backstory, with strong doses of sentiment. And suspects who were eliminated in the initial film step forward as plausible culprits in this one.

Just as important, there’s an added structural premise that gives the new entry an acknowledged affinity with Back to the Future II and other forking-path tales. To top things off, the second installment supplies a revised version of the outcome of the first one.

Although the sequel isn’t thriving [16] at the box office, perhaps there will be a third entry [17].

I have the third movie and I have already pitched it to Blumhouse. Everybody is ready to go again if this movie does well. I keep shifting the tone, genre a little bit. The third movie I know is going to be a little different. It’s going to be really bonkers and really fun.

Bonkers or not, having another version would show that the Variorum never sleeps.

 

Two last points. First, a film I discussed last time [3], Confession (1937), internalizes the variorum impulse in a milder way, by replaying a key scene from a different character’s viewpoint. More unusually, Confession is a remarkably close remake of a German film, Mazurka, and thus constitutes a homegrown variation on the original.

Secondly, why study the variorum principle? Hume points out:

To insist on analyzing the famous writers and plays in isolation is a mistake: much may be learned by viewing them as they originally appeared–variably successful in the midst of a prolific, unstable, and rapidly changing theatre world.

So one argument is that we can best understand and appreciate masterful filmmaking against the background of normal practice. That seems right to me, but I think there are other good reasons to ask these questions.

For one thing, through bulk viewing of a lot of films, you may discover accomplished works. Many well-regarded films have gained their renown through accidents of release and critical reception. (His Girl Friday [18] is one such [19].) Good films lurk in many crevices of film history.

I also think that the norms are of interest in themselves. They can show that craft practices harbor more variety than we sometimes think. Studying norms can also reveal offbeat possibilities that are sketched for future development. In Reinventing, I sometimes point to films, either obscure or awkwardly constructed or both, which anticipate trends to come. One example would be the strange, time-shifting exercise Repeat Performance (1947). It’s a Forties counterpart to Dangerous Corner [20] (1934), a two-path plot looking toward more elaborate forking-path storytelling. It shows as well that rather unusual options can float around the edges of the variorum.


My quotations from Rob Hume come from correspondence and The Development of English Drama in the Late Seventeenth Century, pp. ix and 129. Thanks to Rob for sharing the backstory of the book’s composition. Readers interested in his method can learn much more about it in his later study Reconstructing Contexts: The Aims and Principles of Archaeo-Historicism [21] (Oxford, 1999).

This entry relies on a distinction among a film’s story world, its plot structure, and its narration. The idea is explained in this essay [22]and applied to a single film in a blog entry on The Wolf of Wall Street [23]. Plots with loops and forking paths are connected with the idea that “form is the new content” in films from the 1990s and after. I try to chart that ecosystem in The Way Hollywood Tells It: Story and Style in Modern Movies. See also this entry [24] for a quick summary of early examples of multiple-draft plotting. For more on the virtues of Charlie Chan and Mr. Moto, go here [25].

The following errors in the hardcover version of Reinventing Hollywood are corrected in the paperback.

p. 9: 12 lines from bottom: “had became” should be “had become”. Oops.
p. 93: Last sentence of second full paragraph: “The Killers (1956)” should be “The Killing (1956)”. Arrgh. Elsewhere on this site [26] I discuss Kubrick’s heist film at some length.
p. 169: last two lines of second full paragraph: Weekend at the Waldorf  should be Week-End at the Waldorf.
p. 334: first sentence of third full paragraph: “over two hours” should be “about one hundred minutes.” Doggone.

We couldn’t correct this slip, though: p. 524: two endnotes, nos. 30 and 33 citing “New Trend in the Horror Pix” should cite it as “New Trend in Horror Pix.”

Whenever I find goofs like these, I take comfort in this remark by Stephen Sondheim:

Having spent decades of proofing both music and lyrics, I now surrender to the inevitability that no matter how many times you reread what you’ve written, you fail to spot all the typos and oversights.

Sondheim adds, a little snidely, “As do your editors,” but that’s a bridge too far for me. Instead I thank the blameless Rodney Powell, Melinda Kennedy, Kelly Finefrock-Creed, Maggie Hivnor-Labarbera, and Garrett P. Kiely at the University of Chicago Press for all their help in shepherding Reinventing Hollywood into print.

[27]

Happy Death Day (2017).