The three kinds of surprise

In real life, most of us would be happy to deal with fewer surprises, but in fiction, they’re a delight. Or at least movies and television would like to believe. In practice, twist endings and plot developments that arrive out of left field can be exhausting and a little annoying, if they emerge less out of the logic of the story than from a mechanical decision to jerk us around. I’ve noted before that our obsession with big twists can easily backfire: if we’re conditioned to expect a major surprise, it prevents us from engaging with the narrative as it unfolds, since we’re constantly questioning every detail. (In many cases, the mere knowledge that there is a twist counts as a spoiler in itself.) And Hitchcock was smart enough to know that suspense is often preferable to surprise, which is why he restructured the plot of Vertigo to place its big reveal much earlier than it occurs in the original novel. Writers are anxious to prevent the audience from getting ahead of the story for even a second, but you can also generate a lot of tension if viewers can guess what might be coming just slightly before the characters do. Striking that balance requires intelligence and sensitivity, and it’s easier, in general, to just keep throwing curveballs, as shows like 24 did until it became a cliché.

Still, a good surprise can be enormously satisfying. If we start from first principles, building on the concept of the unexpected, we end up with three different categories:

1. When something happens that we don’t expect.
2. When we expect something to happen, but something else happens instead.
3. When we expect something to happen, but nothing happens.

And it’s easy to come up with canonical examples of all three. For the first, you can’t do much better than the shower scene in Psycho; for the second, you can point to something like the famous fake-out in The Silence of the Lambs, in which the intercutting of two scenes misleads us into thinking that an assault team is closing in on Buffalo Bill, when Clarice is really wandering into danger on her own; and for the third, you have the scene in The Cabin in the Woods when one of the characters is dared to make out with the wolf’s head on the wall, causing us to brace ourselves for a shock that never comes. And these examples work so elegantly because they use our knowledge of the medium against us. We “know” that the protagonist won’t be killed halfway through; we “know” that intercutting implies convergence; and we “know” when to be wary of a jump scare. And none of these surprises would be nearly as effective for a viewer—if one even exists—who could approach the story in complete naiveté.

But not every surprise is equally rewarding. A totally unexpected plot development can come dangerously close—like the rain of frogs in Magnolia—to feeling like a gimmick. The example I’ve cited from The Silence of the Lambs works beautifully on first viewing, but over time, it starts to seem more like a cheat. And there’s a fine line between deliberately setting up a plot thread without paying it off and simply abandoning it. I got to thinking about this after finishing the miniseries Fargo, which I loved, but which also has a way of picking up and dropping story points almost absentmindedly. In a long interview with The A.V. Club, showrunner Noah Hawley tries to explain his thought process, with a few small spoilers:

Okay, Gus is going to arrest Malvo in episode four, and he’s going to call Molly to tell her to come, but of course, she doesn’t get to go because her boss goes. What you want is the scene of Molly and Malvo, but you’re not getting it…

In episode ten when Gus tells her to stay put, and she just can’t, and she gets her keys and goes to the car and drives toward Lester, we are now expecting a certain event to happen. Therefore, when that doesn’t happen, there’s the unpredictable nature of what’s going to happen, and you’re coming into it with an assumption…

By giving Russell that handcuff key, people were going to expect him to be out there for the last two episodes and play some kind of role in the end game, which is never a bad thing, to set some expectations [that don’t pay off].

Fargo is an interesting test case because it positions itself, like the original movie, as based on true events, when in fact it’s totally fictional. In theory, this frees it up to indulge in loose ends, coincidences, and lack of conventional climaxes, since that’s what real life is like. But as much as I enjoyed Fargo, I’m not sure I really buy it. In many respects, the show is obsessively stylized and designed; it never really feels like a story that could take place anywhere but in the Coenverse. And there are times when Hawley seems to protest too much, pointing to the lack of a payoff as a subversion when it’s really more a matter of not following through. The test, as always, is a practical one. If the scene that the audience is denied is potentially more interesting than what actually happens, it’s worth asking if the writers are being honest with themselves: after all, it’s relatively easy to set up a situation and stop, while avoiding the hard work that comes with its resolution. A surprise can’t just be there to frustrate our expectations; it needs to top them, or to give us a development that we never knew we wanted. It’s hard to do this even once, and even harder to do it consistently. But if the element of surprise is here to stay—and it doesn’t seem to be going anywhere—then it should surprise us, above all else, with how good it is.