The last word in films

“But what if ... “

For anyone who’s ever been dissatisfied with a movie ending, those three little words now appearing on 840 theater screens offer a tantalizing rush of mixed emotions.

They’re attached to “28 Days Later,” the successful and unusually intelligent British horror movie, which lately began offering two endings for the price of one. First, you watch the original version, which contains an ending many fans felt was too cheery for a gory tale about the end of British civilization. Then, after the credits roll, those words appear in the lower right-hand corner of the black screen, and suddenly you’re watching a new four-minute ending that is decidedly bleaker but more emotionally in sync with the story.

The second ending, introduced July 25 as “28 Days Later” began its second month of U.S. screening, has been dismissed as a promotional stunt and an abdication of artistic decision-making. It has been praised as a romantic return to 19th century literary tradition, when Dickens penned two endings to “Great Expectations,” and newspapers asked readers to offer their opinions on how to end serialized novels.

But the two-endings gimmick perhaps is best appreciated as a window onto the film business’ chronic inability to end a movie right.

“28 Days Later” appears to be the first film to give movie theatergoers the make-up-your-own-mind experience of multiple endings often offered on DVDs. Yet it is part of a long Hollywood tradition in which clashing pressures of creativity, collaboration and commerce twist and often weaken movies’ last scenes.

Chili Palmer, the cocky mobster-turned-producer, summed this up in the last line of Elmore Leonard’s novel “Get Shorty”: “[Expletive] endings, man,” says a chastened Chili. “They weren’t as easy as they looked.”

What is easy is classifying the many wretched endings littering the cineplexes: There is the predictable ending (virtually any teen movie), where the nerd is redeemed. The uplifting ending that turns a complex story into a happy triumph because the producers got spooked by test-marketing data. The trick ending that leaves you furious because the story never gave you the clues to anticipate it. The loophole ending that lets the rightly defeated bad guy survive to fight another day (read: sequel). The layered ending, in which conclusion after conclusion washes over the impatient audience.

By contrast, think about how seldom you’ve watched a movie end perfectly. Think about “The Sixth Sense” and how beautifully the pieces fell into place once you staggered out of the theater. Think about how “The Usual Suspects” did the same thing as Kevin Spacey virtually morphed into the evil suspect in the final seconds. Think about the proudly ambiguous ending of “Midnight Cowboy”: the way a newly resolute Jon Voight cradles a dead Dustin Hoffman as their bus heads toward what was supposed to be a new life. Think about the last line of the comedy “Some Like It Hot,” when a smitten Joe E. Brown is informed the woman he loves is Jack Lemmon and responds cheerfully, “Nobody’s perfect.” Think about the wordless image of an endless line of cars flowing toward that mystic baseball diamond in “Field of Dreams.”

As Chili Palmer learned, nailing the ending -- whether it’s the third act, the final moments or simply the last words -- is the hardest part of telling any story.

“It’s easier to set up a story than resolve it,” says David Howard, a screenwriter, script doctor, founding director of USC’s graduate screenwriting program and author of two books on writing. “If you set the ball rolling the wrong way, there’s nothing you can do at the end. So often when something needs to be repaired, where you have to start looking is the first act -- Who are the characters? Were they given enough room to change? Was the right subject chosen to come to a satisfying resolution?”

Another screenwriter (“Top Gun,” “Legal Eagles”) and script doctor, Jack Epps, said the trick lies in satisfying two competing demands: The ending has to be surprising -- “We can’t see it coming, we like a good twist” -- but it has to make sense by growing out of the character of the film.

“American Beauty,” for example, stunned viewers when Kevin Spacey’s next-door neighbor murdered him just as Spacey’s character came to terms with his middle-class life. But in the context of the film -- a searing portrait of an American family -- that ending fit, Epps said. “It was shocking, but it felt right.”

This kind of craftsmanship is hard enough for a lowly screenwriter sitting at a computer at 2 a.m. But the way movies are made and sold adds layers of contradictory pressures -- much of it directed at the ending, which is regarded as a key factor in that all-important marketing tool: word-of-mouth buzz. The director’s vision often alters the screenplay. A studio with a particular demographic in mind may ask for -- or order -- additional scenes featuring a specific character, indirectly rendering the ending less coherent. Test-marketing, a staple of Hollywood filmmaking since the ‘30s that has been used increasingly during the past two decades, may show a perceived audience hunger for a less harsh or more redemptive ending.

“It’s a trauma that just about every writer who embarks on a screenplay understands he is going to face,” said Penn State film historian Kevin Hagopian. “You have a lot of cooks who want to get into the conclusion.”

Screenwriter Ira Heffler knew this and was prepared to fight it. There are thousands of Hefflers in Hollywood, people who have sold or optioned scripts but have yet to get an actual movie into theaters. Heffler and a writing partner were working on a script about a guy with no love life who starts interviewing the women who dumped him. Heffler vowed that the story would end realistically -- without a Hollywood ending. “I told my partner I didn’t want [the protagonist] walking off with the love of his life.... He has to learn and grow, to get it.” Then he sold the script to a small independent venture and was told almost immediately: “Change the ending. We want to walk out of the theater feeling good.” Heffler started to fight but was silenced by his agent -- something he’s grateful for today. After all, this film actually wound up getting shot (after other writers were brought in). Yet five years later, the ending remains a bugaboo. Even as his producers seek a distribution deal, director Chris Hall says they are still debating whether to abandon the happy ending and return to Heffler’s quirkier concept.

The struggle in endings, says Hall, a 31-year-old USC film school grad, is often between “the artist and the commercialist in you.”

That was why, for example, Glenn Close’s character died at the hands of Michael Douglas’ character’s wife, rather than by suicide, as originally shot, in “Fatal Attraction.” Test audiences felt the suicide ending ignored their need for Close’s character to die a vengeful death. The public’s perceived desire for a happy ending was also why “Blade Runner,” against director Ridley Scott’s wishes, was given a syrupy conclusion in which the hero’s love interest, one of the genetically engineered “replicants” who lived only four years, somehow wound up with a limitless life span. Eleven years later, Warner Brothers theatrically released Scott’s original version to critical acclaim.

Defenders of test data say it can work when subtleties are respected. In 2001’s “Legally Blonde,” test audience reactions convinced filmmakers to change the ending twice, according to one of the producers, Marc Platt.

The original girl-gets-boy, romantic-comedy ending didn’t let the audience laugh enough -- test audiences cared more about the comedy than the romance. So comedic bits were added, but subsequent test audiences found that ending unsatisfying too, Platt said.

What filmmakers finally realized, he said, was that the audiences had become caught up in the growth of Reese Witherspoon’s spunky character, who improbably graduates from Harvard Law School. “The film that emerged was a woman’s journey, not a romantic comedy,” he said, so the ending was reshot with Witherspoon giving the class graduation speech and incorporating other characters.

“Nobody has a monopoly on good ideas,” Platt said. “What often occurs [in audience testing] is that elements of the film you thought were funny turn out to the audience not to be. Scenes that you thought were boring turn out to be hilarious.”

Producer Mark Johnson (“Rain Man,” “The Rookie,” “Good Morning, Vietnam”) remembers one piece of test-audience ending advice he considered dubious. “On ‘Rain Man,’ a lot of people wanted Dustin Hoffman to stay with [his idiot savant character’s flashy brother] Tom Cruise at the end of the movie.” That would have been happier but nonsensical. “If we had given it to them, I am convinced they wouldn’t have liked it.”

Great endings, like great stories as a whole, grow out of individual inspiration, Johnson said, recalling the way the first “Godfather” movie ended with a door closing on Michael Corleone as his hand is kissed by Mafia underling Clemenza. “That’s not something that comes out of a meeting of producers and studio executives. If you gave that [ending] to a committee or the development process to decide, they’d say you can’t end on such a dark, ambivalent moment; and yet it’s perfect for that movie. There’d probably be a compulsion to make it bigger and more dramatic, something more than less. And it’s a wonderfully dark summation of the whole story.”

Once they get beyond a few universally acclaimed endings -- the courage of “Casablanca,” the freeze-frame shot of Francois Truffaut’s “The 400 Blows” -- film buffs can disagree on almost any ending: Is that freeze-frame ending of “Thelma & Louise” a brilliant way of reflecting on the women’s rebellion, or a cop-out? Did Steven Spielberg really need to wrap contemporary baby-boomer beginnings and endings around “Saving Private Ryan”? Couldn’t he have just ended it where the real WWII story stopped? Was “Fight Club’s” ending chilling or unfathomable? Was the fantasy sequence at the end of Spike Lee’s “The 25th Hour” a slice of misleading self-indulgence or a homage to a literary technique in which we experience a character’s vision of what might have been, only to find out it was not real? Was the last act of screenwriter Charlie Kaufman’s “Adaptation” a cogent satire on Hollywood endings or the real-life last gasp of a writer bereft of more nuanced ideas?

Execution often separates two conceptually similar endings. Walking out of “The Sixth Sense,” you could reassemble the story and see just where the film warned you that the protagonist, Bruce Willis, was unknowingly dead. “Basic Instinct,” the Sharon Stone-Michael Douglas murder mystery, ended with the same kind of goose-bump-inducing revelation but fell apart when you began analyzing the way the film prepared the audience.

The makers of “28 Days Later” were torn between giving their audience a respite after a grueling saga (infection turns virtually all Britons into killer zombies) or leaving them with an ending that resonated with the same hopelessness.

(Spoiler alert: If you don’t want to know the endings, stop reading now.)

Screenwriter Alex Garland and director Danny Boyle (“Trainspotting”) started with an ending in which three survivors appeared to be on the verge of being rescued when a jet passed overhead. But the filmmakers changed their mind and, in the first cut of the film, used an ending in which the only male survivor died and his companions -- a woman and a teenager -- headed into the unknown. Then the filmmakers changed their minds again, convinced that, as Boyle said, “We felt we can’t do this to people” and went back to the upbeat ending. Test audiences reacted better to that one.

The film debuted in Britain last fall. In May, when a British DVD was released, the darker ending was included as an alternate, and many fans began arguing over Internet sites that this ending was more honest. Fox Searchlight, spotting a marketing opportunity for a surprise hit that had already grossed $30 million in its first month, added the second ending.

On a recent Monday morning, only five people made their way to a Third Street Promenade theater to watch an 11:15 showing of “28 Days Later.” One of them, 16-year-old Santa Monica High School student Nicole Hayashida, said she liked the upbeat ending better “because after a whole movie full of death, it gives you hope.” She found the presence of the second, downbeat ending disconcerting.

“It was strange,” she said, “because usually when people make movies, they know how to end them.”

(But what if the story did not end there? In the belief that this story will enjoy better word-of-mouth if it ends with a dollop of humor, rather than the unintended irony of Nicole’s quote, we offer an alternate ending):

“It was strange,” she said, “because usually when people make movies, they know how to end them.”

That’s what Chili Palmer used to think.

(The end)