Written by Bill Mesce
So, what’s wrong with the three-act structure?
I’d go as far as to say the three-act structure may be the most writer-friendly form of narrative architecture there is. It’s logical, common sensical, it has a beautiful ABC elemental simplicity. Rom coms, sci fi, crime thriller…the genre doesn’t matter; it works. Writing for the big screen? Little screen? Stage, Saturday Night Live skits, radio dramas? The medium doesn’t matter; it works.
It works because it’s a natural way to tell a story: “A guy goes into a bar (Act I), says to the bartender, ‘My wife left me today’ (Act II),” etc.
It’s so basic a form that much of the blather about three-act structure used to fill out books on the topic is a lot of padding around this simple formula:
Act I: the set-up i.e. the beginning
Act II: development i.e. the middle
Act II: the conclusion i.e. the end.
Author/screenwriter William Price Fox, who’d been my undergraduate creative writing and screenwriting instructor in those days before “three-act structure” became a commonly tossed-around term, used to describe it to us this way:
“You set up two guys around a tree stump; one guy chases the other guy around the tree stump for a while; the one guy catches the other guy, or he gives up.”
So why is three-act structure on my screenwriting myth list?
Like the nice kid new to the neighborhood who starts running with the thug-types always hanging around the corner bar, it fell in with bad company. Over the last 20-30 years, three-act structure has gone from a writer-friendly tool into something carved deep into granite tablets as an inflexible, only-way-to-go commandment. The present-day concept of the model has become so rigid and restrictive that development execs will even specify at about what page each plot “turn” marking off the acts should take place: “Ok, Bill, around page 30 we have to move into our second act…”
Says who? Was there some research I’m unaware of that says the human brain automatically loses interest if there’s not a major plot turn a half-hour into a movie? If it comes earlier? Later?
I suspect one of the reasons script development people, agents, a young generation of screenwriters, etc. have glommed on to this as a commandment is “three-act structure” has the sound of a long, respectable pedigree, something reaching back to the earliest days of theater: the first form of script writing.
Well, it doesn’t.
Elizabethan drama, for instance (think Shakespeare), has a five-act structure. Go back to the ancient Greeks who birthed this first form of mass entertainment, and their stage works don’t have acts. Instead, they have 10-12 sections, alternating between performers carrying out the plot and a chorus providing the audience with commentary, character insights, back story and other odds and ends. Many, if not most, modern-day plays are written in two acts (although three-act sticklers would make the case that the first act is really two acts jammed together). And I’m not even going to get into the category of one-act and experimental plays.
The point here being that if theater – which presumably gives us the concept of the three-act story – hardly holds any kind of blind allegiance to the structure, why should film?
What this dogmatic commitment to the concept has done has been to take this writer-friendly tool and twist it into a story-killer by infecting the form with the storyteller’s worst enemy: predictability. By banging the structure into an inflexible form, anybody who has sat through just a few mainstream releases can already sense plot turns coming. In some forms that are already fighting an inherent form of predictability (like, say, rom coms and superhero flicks), the infection becomes lethal.
Take one of this summer’s contributions to the superhero category, Ant-Man (screenplay by Edgar Wright, Joe Cornish, Adam McKay and the movie’s star, Paul Rudd). Structurally, there’s not that much difference between Ant Man and Iron Man (2008, screenplay by Mark Fergus, Hawk Ostby, Art Marcum, Matt Holloway), the first Spider-Man (2002, screenplay by David Koepp), or even the first Spider-Man reboot, The Amazing Spider-Man (2012, screenplay by James Vanderbilt, Alvin Sargent, Steve Kloves). The execution gives them different flavors (Robert Downey, Jr.’s breezy, witty Tony Stark in Iron Man; the geeky charm of Tobey Maguire in Spider-Man; director Marc Webb bringing a sweet young-love awkwardness to Amazing), but if you strip all four of those movies down to a plot schematic (social oddball comes into possession of superpowers, wrestles with responsibility of having superpowers, assumes that responsibility and faces down nemesis who manages to put one of the hero’s loved ones – Ant-Man’s daughter, Tony Stark’s secretary Pepper Potts, Spidey’s crushes Mary Jane Watson and Gwen Stacey respectively — in danger), it’s pretty much the same thing. The reason Ant-Man is so terminally bland is because it’s so terminally bland; there’s not enough flavor to distract you from the fact you’ve seen this story before. A lot. That you knew how it was going to play out when you saw the coming attractions.
For a storyteller, that’s Death On Wheels.
I became aware of how fatal that kind of predictability (combined with familiarity) could be stuck on an airliner where the in-flight movie was the gooey 1996 romance Bed of Roses with writer/director Michael Goldenberg stirring in the goo. Christian Slater is a shy but talented florist, Mary Stuart Masterson a hard-driving career woman, and as mismatched as they are they still click (non-surprise Number One), they have a misunderstanding and almost break up (#2), they overcome their misunderstanding and wind up living – you guessed it – happily ever after (#3). Maybe if you were a Martian (and a pretty oblivious Martian at that) who’d never seen an American romance flick made in the previous 20 years, any of these see-it-a-mile-away plot turns might’ve surprised you. Otherwise, instead of being pulled along by the story, you’d be ahead of it, glancing at your watch while you wait for the and-there-it-is moments you knew were coming. Storytelling Death.
Some screenwriters have gotten around that predictability by scrambling the chronology of their stories. A favorite of mine is (500) Days of Summer (2009, written by Scott Neustadter, Michael H. Weber). Neustadter/Weber reshuffle the 500 days of the relationship between greeting card writer Joseph Gordon-Levitt and free spirited Zooey Deschanel (the Summer of the title). For instance, the movie skips ahead to the day when Gordon-Levitt finally sleeps with Deschanel. Neustadter/Weber and director Marc Webb pull out all the stops to capture Gordon-Levitt’s exhilaration. Not only is he singing and dancing his way through the city, but passers-by around him join in, and so do a couple of animated birds. Then we slam-cut to several hundred days in the future and see that same Gordon-Levitt standing in an elevator, disheveled, depressed, dejected, and whatever other de-fill-in-the-blank you can come up with. The conjunction of the two impressions – delirious joy/spiritual destruction – is funny as hell, but that gap of several hundred days creates mystery, suspense, questions. That’s the definition of audience engagement. As the movie pinballs back and forth through the 500 days, the audience has to mentally assemble the chronology and emotional arcs, none of which are complete until the last pieces fall into place.
Betrayal (1983), Harold Pinter’s adaptation of his play, tells the story of a marriage/friendship-destroying affair in reverse order. Christopher Nolan’s screenplay for his thriller Memento (2000) told his story of an amnesia-afflicted man trying to avenge the murder of his wife through two narratives: one moving forward in time, the other in reverse. And then there’s Quentin Tarantino’s scrambled narrative for Pulp Fiction (1994).
Tarantino was actually borrowing a structure Stanley Kubrick used in one of his early features, the 1956 crime thriller The Killing. Kubrick regularly halts the forward narrative of his heist flick, then retells it from another character’s perspective. If you were to break the movie up into its various narrative lines, you’d wind up with several mini-movies, each with its own three-act arc, but Kubrick’s artful jumbling regularly has us witnessing what-the-hell-happened? moments, then answering them by rewinding the narrative and showing us that part of the story from another perspective. Like (500) Days, Betrayal, Pulp Fiction etc., Kubrick pulls the audience in deeper by making viewers have to work out how all the pieces of the puzzle come together.
These flicks are also a testament to the hardy nature of the three-act structure because, at heart, they don’t violate the model as much as they stretch, twist, compound, and toy with it.
What they also show is there’s no emphatic rule about when demarcating plot turns need to happen. If you want to look at one movie that puts a stake through the heart of that bit of arbitrary dogma, check out Frank Capra’s Christmas classic, It’s a Wonderful Life (1946, screenplay by Capra, Frances Goodrich, Albert Hackett, Jo Swerling, and an uncredited Michael Wilson).
For those of you who’ve managed to miss the annual holiday airing of this eye-dampener:
Prayers reach Heaven on behalf of George Bailey who, faced with financial ruin, is ready to throw himself off a bridge. His guardian angel is briefed on what put poor, forlorn George in such a spot. The movie then tells – at great length – George Bailey’s story in a series of episodes ranging from boyhood to the present. Always a boy-then-man of high ambitions and dreams, George Bailey’s life is one frustration after another, necessity and then the responsibilities of family dismantling his once grand vision of his own future piece by piece until George is left with a family he can barely support, a financial hole threatening to put him in jail, and a lifetime of lost and regretted opportunities.
Wonderful Life’s first act takes up a little more than 75% of the story. You read that right: three quarters of the movie is the set-up. Act II – where George is granted his wish that he’d never been born and sees the crap-hole his hometown would’ve become without him – takes up about 15-20%, and the quick dash that’s Act III — George realizes he’s been so focused on the negatives of his life that he’s missed the golden positives and wishes the world be put back the way it was, including his financial problems – gets the rest.
As out of whack as this construction is by contemporary standards, Wonderful Life only works because of that meticulous construction of the life of George Bailey, the audience living George’s mounting disappointments along with him. Imagine It’s a Wonderful Life with the events of the first act compressed into the first 20-30 pages, and the second act padded out to about the same. Do that and what you don’t wind up with is a movie that has damned near everybody watching it on TV reaching for the tissue box.
Wonderful Life’s proportions may be off the modern-day scale, but it’s still a classically-structured three-act story. There are, however, movies in which storytellers just flat-out disregard form.
The classic anti-Western, The Wild Bunch (1969, screenplay by director Sam Peckinpah and Walon Green) begins its 145-minute run with an approximately 15-minute section in which the titular bunch – a group of ruthless outlaws – ride into a Texas border town disguised as soldiers to rob a railroad office, are ambushed by waiting bounty hunters, then fight their way out in what is still one of the most brutally violent scenes in mainstream cinema. However, the forward movement of the movie doesn’t begin until sometime after this sequence (about 50-minutes in) when the robber band’s survivors drift into a Mexican town and hook up with a local warlord. This prologue, like the irrelevant-to-the-main-plot Omaha Beach sequence in Saving Private Ryan (1998, screenplay by Robert Rodat), is about setting a tone, a sense of the mindset of the principals, a sense of dread now that we know what’s possible, what these characters are capable of, what mental scars may mark them.
You could eliminate these sequences and the proper plot of each movie would still be intact; each would still have a beginning, middle, and end. Don’t believe me? Try coming in after those sequences. You’ve still got a movie. Just not the classics we know.
One structure three-act fanatics have labeled anathema, blasphemy, taboo is the “episodic” structure. You can see episodic films at work – and working quite splendidly, thank you – during the heyday of the Hollywood epic.
First, you’ve got to readjust your head a little. These days, when every summer weekend sees some new $200 million FX extravaganza roll into multiplexes, the idea of an “event” film might be meaningless. But there was a time in pre-CGI antiquity when blockbusters like Gone with the Wind (1939), The 10 Commandments (1956), The Bridge on the River Kwai (1957), Ben-Hur (1959), The Longest Day (1962), The Big Country (1958), The Sand Pebbles (1966), Lawrence of Arabia (1962), etc. were huge events if, for no other reason, in those days they were so damned hard to make. Maybe only a handful of these behemoths rolled into theaters in a year, playing, at first, in only the biggest movie houses in the biggest cities, complete with musical prologues, intermissions, and souvenir programs. Like I said: huge!
They were often based on equally mammoth books, and it was the intent of their makers to, in their film adaptations, try to capture the same texture and dramatic gravitas of their source novels. Big novels tend to be episodic and so did these epics.
Three-act police, trying to force-fit their structure, will say the middle hunk of these films (and it was a big hunk; the running times of these gargantuas could run from 2-1/2 hours to 3-4 hours) is actually a second act and in some cases that’s true. In a lot of cases, though, that’s pure self-justifying crap.
Let’s look at the 182-minute The Sand Pebbles (adapted from Richard McKenna’s novel by Robert Anderson). The set-up is relatively simple and quickly laid out: Steve McQueen plays a U.S. sailor assigned to a river gunboat in 1920s China only interested in running his little engine room without interference. But over the course of the film McQueen’s character lives through a series of mini-arcs – episodes with their own three-act structure – that do not comprise a traditional second act. For instance, McQueen gets into a feud with the Chinese coolie in charge of the engine-room coolies, the Chinese dies in an accident, McQueen trains a new coolie with whom he develops a friendship, the coolie has to fight one of the biggest bullies on the gunboat to earn his place in the engine room, then later he’s sent ashore by the ship’s chief coolie in a demonstration of power that results in McQueen’s friend being captured and tortured by Chinese communists and in McQueen having to mercy-kill his friend. At that point, the movie is far from over. There will be other incidents (a friend of McQueen’s marries a Chinese woman, dies from illness and she is murdered by the communists; a humiliating march through a Chinese city; and more).
The Sand Pebbles is not a plot-driven story. It’s a character piece, and we can see how each of these episodes begins to alter the isolationist McQueen character into someone who ultimately has to take a rebellious stand. The length of the movie, the broad tapestry formed by these independent episodes immerse us in McQueen’s world in a way shorter, more plot-driven films usually don’t (and sometimes can’t).
Or consider The Godfather (1972). The opening wedding scene has little to do with moving the film forward. Yes, it introduces us to the principal characters, but spends even more time in a static basking in a family extravaganza that, in itself, tells us something about these characters and the world they live in (outside in bright sunlight, a party celebrating life; inside in shadow, discussions of violence and moral corruption). Then follows the Johnny Fontaine episode (climaxing in the infamous horse’s head scene) which has nothing to do with the central plot of the movie. It is, in fact, a digression, but one that shows us that grandfatherly Don Corleone (Marlon Brando) is, in fact, capable of incredible brutality.
The movie doesn’t gain any true forward momentum until Corleone turns down the offer of getting into the drug business with Al Lettieri’s Solozzo. By that time, we’re somewhere around a half-hour or better into the movie.
And then there’s The Longest Day, a sweeping recreation of the Normandy invasion during WW II. Like the Cornelius Ryan book it’s based on (adapted by Ryan, Romain Gary, James Jones, David Bursall, Jack Seddon), Day is less a story than a mosaic comprised of dozens of independent incidents and mini-arcs. Day is the definition of “episodic,” and a perfect illustration of why “episodic” can work…and work beautifully.
Or Citizen Kane (1941). 12 Angry Men (1957). Odd Man Out (1947). The first third of The Magnificent Seven (1961). Give me some time, I’ll find you more, but next time you hear a development guy or agent or writing guru throw up his/her hands in dismay sighing, “This is too episodic,” keep in mind this prestigious company.
You can’t talk about three-act exemptions without talking about Stanley Kubrick, because no filmmaker working in the commercial mainstream pushed further outside movie-making conventions than Stan the Man. We’ve already mentioned The Killing, but that movie seems practically traditional in structure compared to 2001: A Space Odyssey (1968). One of my undergrad film teachers, the esteemed academic and one-time host of the PBS series, The Cinematic Eye, Dr. Benjamin Dunlap, said the way to understand 2001 was to look at is as a “tone poem” (not unlike Richard Strauss’ “Also sprach Zarathustra,” which became the film’s signature piece of music) rather than as a story. The movie (which Kubrick co-wrote with sci fi maven Arthur C. Clarke) is divided into four barely connected parts: “The Dawn of Mankind” taking place millions of years in prehistory, then the “present” discussing the discovery of a mysterious monolith on the moon, then 18 months hence and a space mission to Jupiter, and finally a psychedelic extravaganza representing a transition from human to “star child.” While there is a direction to the story, the movie is, for all intents and purposes, close to plotless, lacks any substantive characters, and its sparse dialogue is, by design, forgettable. Kubrick wanted, in his own words, 2001 to be a“…a non-verbal experience…” and it’s not only non-verbal, but a non-narrative, non-traditional experience as well.
Kubrick went for something similar in his 1987 Vietnam War epic Full Metal Jacket which was also comprised of barely-connected parts and had no particular narrative through-line or character arcs. According to Kubrick’s Full Metal collaborator, Michael Herr, Kubrick was not interested in telling a war story as much as presenting an observation of the “phenomenon” of war.
So, not to belabor the point any more than I already have, my point is not that there’s anything wrong with three-act structure, but that it’s not the only game in town. Three-act structure is a hardy, strong house; four stout walls holding up a strong roof. But Frank Lloyd Wright showed us is that’s not the only to build a house.