I tried several times to call her, but after the first call, she wouldn’t come to the phone any longer. I also sent flowers but with no luck. The smell of the flowers only made me sicker. The headaches got worse. I think I got stomach cancer.
The first time I watched Taxi Driver, that brief snippet of narration caught me off guard. I sincerely believed I understood Travis Bickle, his desires, his demons—but that sudden, violent swerve in his train of thought forced me to reevaluate everything I thought I knew.
Screenwriting teachers frequently discourage students from abusing voice-over narration—after all, why rely on a character’s words when you can convey so much more through their actions (in other words: “show, don’t tell”)? But Schrader and Scorsese don’t use the device as a crutch to prop up flimsy characterization; on the contrary, Taxi Driver’s often contradictory narration adds a whole new dimension to Travis’ character. His disturbing journal entries neither illuminate his motives nor clarify his behavior; they only deepen the enigma that is Bickle, contributing to the story’s oppressive sense of discomfort and imbalance.
Scorsese utilizes V.O. with similar creative flair throughout his body of work. The near constant narration in Goodfellas, for example, adds to the picture’s machine gun rhythm, complementing the energetic visual flow to create a sort of street gangster poetry. More importantly, it implicates the audience in the onscreen action. When Henry Hill introduces us around a table (“Then there was Jimmy Two Times, who got that nickname because he said everything twice…”) and the assembled mobsters directly address the camera, we become part of “the family,” reinforcing our immersion in this unfamiliar culture—the glamour, the decay, and everything in between.
After Goodfellas, Scorsese continued to twist and dissect the traditional forms of narration. The dueling voice-overs in Casino contain some of his most ambitious experiments. In one pivotal scene, a previously minor character (played by the one and only Frank Vincent) suddenly interrupts the action to share his innermost thoughts, providing a unique perspective of his personal stakes in the conflict (“What could I say? If I had given them the wrong answer, Nicky, Ginger, Ace—all of them could have wound up getting killed.”). More famously, the same character cuts off Joe Pesci’s last bit of narration mid-sentence by brutally beating him with a baseball bat.
My favorite example of Scorsese-style narration, though, occurs in his 1993 adaptation of Edith Wharton’s The Age of Innocence. Throughout this tense costume drama, a disembodied female voice (provided by Joanne Woodward) describes, explains, and otherwise clarifies the myriad customs, mannerisms, and rituals of 1870s New York. At first, it sounds an awful lot like a typical inelegant exposition dump; only gradually does its true purpose become clear:
He guessed himself to have been, for months, the center of countless silently observing eyes and patiently listening ears. He understood that, somehow, the separation between himself and the partner of his guilt had been achieved. And he knew that now the whole tribe had rallied around his wife. He was a prisoner in the center of an armed camp.
The men and women drifting through Scorsese’s vision of the 19th century are far too well-mannered/emotionally-repressed to talk about their feelings—even (especially?) in the privacy of their own thoughts. They need an outside observer to express their passion, their lust, their despair for them. Like Taxi Driver, like Goodfellas, like Casino, the voice-over creates a rich interior landscape that enhances what we see (in the case of Age of Innocence Daniel Day-Lewis’ rigid body language and subtly tortured eyes), rather than merely reiterating it.
More filmmakers should strive to use this much maligned storytelling technique so gracefully.
[Originally written April 1, 2012.]