There has been no American feature story more honored – or over-praised – than “Snow Fall” by the New York Times. I don’t want the key word in that last sentence – over-praised – to detract from the story’s historic achievement. It won a Pulitzer Prize in 2013 for feature writing; it set a standard for multi-media reporting at a time when we were wondering about the viability of that form of storytelling; and it attracted attention from far and wide, lending encouragement that journalism in the digital age has an exciting future.
Cheers to the writer, John Branch, to graphics director Steve Duenes, and to the team that created it.
Much of the original praise for the work was worshipful and, I believe, superficial. The dazzling visual effects were there for all to see and left potential critics, dare I use the term, snow-blind.
“Snow Fall” is many good things, but great storytelling is not one of them. I will argue that the innovative visual elements that brought it fame detract from the power of the narrative.
(Let me say that it is exactly this kind of work – impressive and innovative – that deserves our sharpest critical attention over time, the way that the Pulitzer-Prize winning novel The Goldfinch by Donna Tartt received darts from critics such as Francine Prose and James Woods.)
The criticisms I bring to “Snow Fall” are mine and mine alone, which is not to say that I am a voice in the wilderness. After the critical euphoria that followed publication, a more sober reflection became possible only with the passage of time. That’s why I am weighing in now. It also provides an opportunity to articulate standards and strategies for what constitutes excellence in multimedia storytelling.
I began to hear tough questions about the “Snow Fall” story – its methods and effects – in coffee-houses where writers meet, at narrative conferences, in newsroom and classrooms. It did not come from jealousy or fear, but from an honest desire to know the strengths and limitations of such an approach.
Using the first section of the story as a kind of microcosm – and encouraging you all to read that much, at least – I offer the following critiques:
1. Most of the visual elements violate the essential credo of effective storytelling: Not to give away too much too soon.
2. Time and again the visual elements “step on the narrative.”
3. There is a “kitchen sink” feel to the visual aspects of the story.
4. The first section of text feels out of tune with the second.
5. Most of all, there is no harmony between what I would call the Voice of the story and its Vision.
Let me expand on these, in order:
1. Giving it away: Is “Snow Fall” a “what” narrative or a “how” narrative? It makes a difference. In a “what” narrative, the reader is driven to learn what happens next. In a “how” narrative, the reader may already knows what happened, but is eager to learn how it happened. “Snow Fall” can’t decide. As a result, there is no “engine” in the story, no driving force of curiosity, as in “who will live and who will die.” Early in the story, we are introduced to a video interview of one of the survivors. We see her whole and healthy, speaking from some unknown location, at a physical and temporal distance from the action described in the story. And yet here is the text beside her image: “Saugstad was mummified. She was on her back, her head pointed downhill. Her goggles were off. Her nose ring had been ripped away. She felt the crushing weight of snow on her chest. She could not move her legs. One boot still had a ski attached to it. She could not lift her head because it was locked into the ice.”
This feels like one of a number of instances in which the textual and visual elements are out of sync.
2. Stepping on narrative: The storytelling that begins “Snow Fall” is of high quality. It shows everything you would want from a narrative: intense scenic action, a dramatic setting, characters we might come to care about, and a cataclysmic inciting incident – an avalanche — so powerful that it transforms a landscape and the humans who enter it. The defining effect of narrative is to create a vicarious experience, to inhabit a world, to be there – on the snow and then in the snow and then buried by snow. The problem is that the narrative line is interrupted, time and again, by elements that are marginal to the storytelling. Embedded in the text are tiny icons that signify the visual tools: a video, a slide show. Every time I clicked on one of these, it took me away from the story. Instead of time moving, time was frozen, so to speak. Narratives are, by definition, linear (although they can contain more than one line). And it is possible to create narratives with multi-media components: text, visuals, audio, music and much more. We have a name for these. We call them movies.
3. Kitchen sink: I’m trying to learn the names for all the multi-media effects use in “Snow Fall.” There is a title page on a mountaintop, where the snow seems to be blowing; there is the way a text or image seems to surface from the bottom of the page before you’ve scrolled down to it; there are videos and slide shows; still photographs; a turning aerial view as if seen from a drone. It is the accumulation of visual elements that created the Gee Whiz response from admirers. But why all these elements? Perhaps a little selectivity was in order. It was Miles Davis, the jazz artist, who talked about how long it took to learn which notes to leave out.
4. Out of tune: Most of my argument is predicated on some sort of disharmony between the words and visuals in “Snow Fall,” but there is also a case to be made about dissonance in the text itself. Even without the visual elements, we would find interruptions in the narrative. The second part, titled “Tunnel Creek” feels as if written in a different voice from the opening scenes. This is a predictable and acceptable practice in all forms of journalism. One genre is called the “broken line,” a hybrid of narrative and informational reporting, the kind of news feature story that might begin with a narrative lead and be followed by a nut graph. Books by authors such as John McPhee, who tends to tell stories about people and their passions, move easily between explanatory and storytelling modes.
But I have a confession here. On three separate tries I was not able to navigate easily through the text of the Tunnel Creek section, which describes topography and history, not because the prose is insufficient, but because it felt like such a departure from the vivid style that begins the piece.
5. Voice and Vision: I believe I can summarize my critique in talking about two effects of creative journalism, one which I will call Voice and the other Vision. Each one can be perceived in a story, especially a work created through a multi-media approach. Each one is a collection of choices made by the writer and producer – in collaboration, one would hope.
Let’s start with Voice, and I’ll borrow my definition from Poynter friend and colleague Don Fry: “Voice is the sum of all the choices made by the writer that create the illusion that the writer is speaking off the page directly to the reader.” The reason you can identify – without bylines – the author of columns by Anna Quindlen and Maureen Dowd is all about voice. Even if the topic is the same, they SOUND different.
So what are some of those choices, the writer might make:
— The level of language, from slang to formal.
— The type of narration, from first person to third person.
— The story form, from experimental to conventional.
— The stance from neutral to partisan.
–The language, from plain to metaphorical.
Using Voice as an analogy, let me attempt a parallel definition for Vision. “Vision is that quality created by the sum of all the choices made by the designer or artist, the effect of which is a unified way of seeing, as if we were all looking through the same lens.”
So what are some of the choices the visual journalist might make that would influence vision?:
— The decision to use color or black and white.
— The development of a color palette consistent with the content and purpose of the work.
— The decision to use visual images as decorative, illustrative, or documentary in nature.
— The choice of visual platforms from still images to video to animation.
— The level of visual elements, from popular and commercial to high art.
— The choice of typefaces.
The more that the collaborators can discuss in advance the elements of voice and vision, the more they can channel their moves to fulfill the mission and purpose of the work, the more successful will be the experience of the audience.
Let’s take an example of another New York Times multi-media story that, in my opinion, offers a more successful marriage of Voice and Vision. The work is called Tomato Can Blues, written by Mary Pilon, and tells the story of a marginal cage fighter who fakes his own death because he owes money to drug dealers.
The tone of the piece, and the elements of the prose style, is reflected in the title. In the slang of the fight game a Tomato Can is a journeyman boxer who almost always loses and who takes a lot of punishment, getting beat to a bloody pulp – like the contents of a can of tomatoes. That slang, that grit, that working class lingo seeks its equivalent in the visual elements, which it finds in the comic book/graphic novel illustrations of Attila Futaki, who created them journalistically, based them on “police records, witness accounts, photographs and the reporter’s notes.” There is even an audio element that matches the words and visuals and is recited by Bobby Cannavale, an actor known for his work on gangland shows like Boardwalk Empire.
So there you have it, one work in which Voice and Vision seem slightly out of tune, and another where they work in close harmony. I look forward to the next experiment in multi-media storytelling by the New York Times and other media organizations.
Comments