Double Feature

Film noir buff James Cutting marries psychology and cinema  Film noir buff James Cutting marries psychology and cinema   It was a tough job, but somebody had to do it. Over the course of a year and a half, psychology professor James Cutting and his grad students spent their days and nights analyzing 150 Hollywood […]

Film noir buff James Cutting marries psychology and cinema

Film noir buff James Cutting marries psychology and cinema


It was a tough job, but somebody had to do it. Over the course of a year and a half, psychology professor James Cutting and his grad students spent their days and nights analyzing 150 Hollywood movies—from Cinderella to Sunset Boulevard, Airplane! to Urban Cowboy. Their results, published last winter, exposed an intriguing correlation between modern film editing and the rhythm of the human mind.

Cornell Alumni Magazine: Why is cinema a useful laboratory for human cognition?

James Cutting: Until five or ten years ago, most psychology was interested in perception in the instant. So in the typical experiment, I'd flash you something and you'd respond as quickly as you could, completely disconnected from everything else. There's a lot to be learned that way, but we've done that for a hundred years; there aren't many paradigms for studying how people understand things over long periods of time. And movies are two hours or so. It's kind of the perfect length, otherwise you get into things like diurnal rhythms. Here's a time that people set aside to have a story told to them.

CAM: How did you get onto the idea of exploring the psychology of film editing?

JC: The key insight that I started with is, "Why does a 'cut' make sense?" It's completely unlike anything we evolved to see—and we not only accept it, but it's used to pace our experience in film. So I wanted to understand more about how cuts work in movies. A psychologist friend was interested in patterns of reaction times in experiments; say you have a bunch of letters and you have to say as fast as you can whether it's a word. You see a lot of responses around 400 to 500 milliseconds, and some that are like two and a half seconds. I'd been looking at cuts in films and I realized the patterns looked the same.

CAM: What types of films did you study?

JC: We joined Netflix and chose 150 films—ten in each of fifteen years from 1935 to 2005. The ones since about 1980 were among the highest grossing of the year, and those before that were the ones that the most people had written about on the Internet Movie Database. We chose five genres—action, adventure, comedy, drama, and animation—and set out to divide each movie into its cuts. We spent about twenty hours on each film over the course of a year and a half, and I think we got about 99 percent of the cuts. When we finally had the data, we could figure out the lengths of the shots, so we had a string of analyzable numbers. And the interesting result was that we found that films, particularly since about 1960, have a profile that's not unlike people's attention pattern as manifested in those experiments, which is called a 1/f distribution.

CAM: So you mean the movies catered to people's attention spans?

JC: This is not about attention span; attention span is something you can hold for five minutes, five seconds, or whatever. Its actually your rhythm of attention. Our brains are incredibly complex, and most complex systems run in oscillations or fluctuations. Movies are just about the most absorbing artistic domain that one can imagine—so the idea is that through the pattern of cuts, the external visual stimulation is essentially grabbing onto your brain and driving it in the kind of pattern it enjoys.

CAM: How much variation did you find in the films you studied?

JC: A lot. I was just looking at Mr. Roberts, and it's so random. And The Seven Year Itch—just random. I love film noir, and it turns out that the shot lengths in most of film noir are completely random—they're as far from this pattern as one can imagine. But it changes beginning in the Sixties. The earliest James Bond film we have is Thunderball, and all of the sudden it's pretty close to 1/f, and there are others that do that, but not many. By about the 1990s, more and more films have a pattern closer to this. Back to the Future turns out the be the film that is closest to 1/f that we found in its genre. Star Wars Episode III: The Revenge of the Sith—dreadful movie—has a steeper profile than 1/f; it has actually gone beyond this particular pattern.

CAM: Was there any correlation between how a film conformed to this rhythm and how much money it made?

JC: None. Zero.

CAM: Doesn't that seem odd?

JC: Not to me. The idea that a film is hitting you with this set of patterns, and that pattern is driving your attention, doesn't have to do with the narrative—it has to do with the presentation of the narrative. There's almost zero correlation between this 1/f fluctuation and how much people like it.

CAM: And if a film doesn't conform to this rhythm, it's not a recipe for disaster?

JC: Absolutely not. For example, Goodfellas doesn't come close; dramas tend to be farthest from this pattern. It's not just action films that fit it. There are a few "chick flicks" as well; Pretty Woman almost exactly fits this pattern. But there's enormous variation. Some of the press about this said, "Formula Found for Hollywood Blockbusters!" But we didn't just do blockbusters, and it's not a formula—it's a pattern.

CAM: How has this affected your personal movie viewing experience? Do you think, There goes that 1/f pattern again?

JC: Almost not at all—unless the movie is boring me.

CAM: So what's your favorite film?

JC: When someone asks me that, I tend to say the one I saw most recently. I just love the experience. I've always enjoyed long shots, and I decry the fact that modern movies don't do them. The opening shot in Orson Welles's Touch of Evil or Alfred Hitchcock's Rebecca—some of these older films have such exquisite shots, they take my breath away.

— Beth Saulnier

Join the conversation

Your email address will not be published. Required fields are marked *

All comments must comply with our Commenting Policy.