begin with the end in mind: some thoughts on the snowflake method

“Begin with the end in mind” is, in retrospect, not an original idea; it is Habit 2 of the Seven Habits of Highly Effective People, and as such has been featured on productivity blogs from here to the ever-expanding edge of the universe. The goal of this post, though, isn’t just to articulate this moral, although it bears articulating. The point is to cash out a little bit of what that actually means when you’re trying to write a big, complicated piece of writing.

Writing is like driving at night in the fog. You can only see as far as your headlights, but you can make the whole trip that way
—E. L. Doctorow

I think it’s hard to achieve a satisfying “end” with something you can keep in mind all at once. Our working memory has limited capacity; we can’t mentally manipulate very complex things in their entirety, and masterful stories require complexity. This is OK if we can rely on the chunks we can store—those that we, in the linear act of writing, create—to cohere. But, in my experience, you can’t make the whole trip that way. This isn’t an a priori argument; writers often speak of faith in the process or themselves, and they often, I think, throw many balls in the air in the hope that a few will arc gracefully and be catchable in some interesting position (and the rest will be erased in revisions). I’m no one to criticize anyone who’s successfully created a work of fiction this way, but I don’t think it’s he best way for me.

So there we are. What are the other options?

Randy Ingermanson’s Snowflake Method

I tried this out on THE DANDELION KNIGHT. I couldn’t get all the way through it — actually, I only made it to about step 3. This is partially because it felt like regressing, since I’d already written several thousand words and a partial outline.

“Step 1) Take an hour and write a one-sentence summary of your novel.”

With a novel already in progress and a hazy idea of the end in mind, this was easy:

A locksmith’s daughter loses her memory when she becomes President of Altronne, a city at war with an enemy from within who goes by the name of the Dandelion Knight.

Looking back on it, though, what’s wrong with this sentence?

The problem is expressible grammatically: What’s the verb? “Loses her memory.” (Verb phrase, all right.) Chapter 1 is due for heavy revision, but I feel more than comfortable stating that, when the dust settles, you’ll know that the main character’s lost her memory well before you turn the first page, and that she’s become President of Altronne well before any serious plot happens. (This flaw is the only reason I posted this sentence in the first place.) The embroidery on the end of the sentence makes it seem more substantial, but it’s just modifying Altronne—it goes to setting, not to plot, and plot is what you want in this sentence. Consider Ingermanson’s summary of his own first novel: “A rogue physicist travels back in time to kill the apostle Paul.” That leaves a lot open—you wouldn’t claim you knew the plot on that basis—but it does sketch out a basic arc in the most economical of terms. Subject, motivation, action. Mine has a subject, almost no action, and little motivation.

I could argue, in my own defense, that I already had a bunch of plot laid out, so I didn’t need a sentence to summarize it. This might or might not be true—as I write, there’s still a lot of uncertainty—but it’s beside the point. The point is that not every sentence that plausibly encapsulates your book will serve as a useful guide in expanding it. “Four hobbits find themselves flung to the ends of the earth as an evil force threatens the kingdoms of men and elves” is accurate, but basically passive; again, it’s got a subject but no real action and no motivation. Try “Four hobbits strike out to rid the world of an evil artifact by returning it to the fires in which it was forged.” It’s not genius writing, but it gives them a goal and at least some kind of action.

One of the things revealed by my flawed summary sentence is my problem drawing out a clean arc of action. I have a lot of trouble constructing plots that make sense—I’d written a lot of action, but there was no arc, no sense of progress or buildup, just meandering. For me, anyway, this is one of the casualties of the headlight approach. When I come to a fork in the road, what determines the decision isn’t just my ability to see the fork. I might have to reexamine what came before, even a while before.

Step 2) Take another hour and expand that sentence to a full paragraph describing the story setup, major disasters, and ending of the novel.

This, however, was phenomenally useful. I don’t like Ingermanson’s tendency to think formulaically, but it’s a really helpful antidote to my own tendency to wander. I think I even adopted his optional “three disasters plus an ending” structure, while still thinking about the book as two more or less equal parts: Part 1 had the first two disasters, while Part 2 had the third disaster and an ending.

Of course, if you read the current (unfinished) draft of the book, you’d laugh at how much it differs from the paragraph synopsis. This is supposedly fine, and there’s no reason it shouldn’t be; Ingermanson may be a formulaist, but he’s also a craftsman, and he’s written enough novels to know that you do get good ideas and find useful refactorings during the act of writing. But this does highlight one problem with the snowflake method: The necessity of syncing up all your “design documents” with the current state of the product. It gets worse as you get farther along in the process, with more documents to sync.

Planning to write is not writing. Outlining, researching, talking to people about what you’re doing, none of that is writing. Writing is writing—E. L. Doctorow

Some of what I’m doing here, of course, is arguing that this is, to a meaningful degree, not true. But, to a meaningful degree, it is true—there’s a surprising number of things a writer will do to avoid writing, and Doctorow says this as an admonition against taking credit for reading the Wikipedia entry on snipe hunts when you ought to be writing a scene about a snipe hunt. It’s well to remember that nothing can replace your end product: Words on a page, written for a reader. But that doesn’t mean the simple method of generating candidates for those words at the expense of all other activities is the best way to produce the work. Headlights aren’t any good without a direction. Some people may be dead reckoners; some have to plot their course.

Step 3) For each of your major characters, take an hour and write a one-page summary sheet…

This is a fun step because it’s a lot like character generation in Dungeons and Dragons (speaking of which, RIP Gary Gygax), which everyone knows is the best part. I often throw a new character into my fiction because the main character’s been wandering around in his or her own head too long, and only later do I realize that this new person is here to stay and had better be recognizable and interesting. So this step affords the opportunity to focus on one character at a time and give the thought to that person’s nature, motives, and trajectory that will inform your portrayal of that person throughout the whole book.

The summary of the character’s storyline also forces you to do some useful thinking. A common flaw in my fiction is that the main character is extremely passive: I’ll use her as a lens to examine a situation or a setting, but the character herself doesn’t make any decisions of consequence, instead swept up in the rush of events or the actions of secondary characters. This kind of problem can be relatively unobtrusive if it’s not examined in isolation. Similarly, you can spot a character who’s required by the plot to be dumb in a critical situation, or one whose actions are extremely repetitious, inexplicable, or contradictory. By subordinating the overall plot to the agency of the character, you help yourself get beyond what you thought the book was about and into what a person would do in the situation you devised.

Finally, this is the first step of the snowflake method in which you really start to think about the action at a fine grain. In my paragraph summary, at least, almost all the interactions among characters were abstracted out into the general plot; there wasn’t much consideration of relationships, interactions, intentions, emotions, false beliefs. This is where you can start to plant seeds of the kind I talked about in the last post, and try to tweak personalities and situations to fertilize those seeds.

Post-finally, I suspect Ingermanson’s assertion about editors (and presumably people in general) preferring “character-based fiction” is exactly right. It’s often hard for me to remember that characters are all there are; no one would read a novel about the trees in a forest jockeying for sunlight, or about plate tectonics, no matter how inventive and lush and exact.

Steps 4 and onward…

This is where I really start marching in step with Doctorow’s “writing is writing” beat. I am not (yet?) a starving writer; I’m a graduate student with limited writing time, and I need to feel some sense of progress, which inevitably means writing some of the draft, or else I’ll give up or go crazy. What this means is that I haven’t yet mustered the discipline to pour as much effort as these steps seem to require into something that is not, itself, the novel. If I’m writing a novel, I want to write the novel. I spent maybe an afternoon on Steps 1-3; Step 4 could end up taking most of my writing time for a month. If I had more time to devote to this, or if I could divide my writing time between (e.g.) short stories and novel-plotting—or even if I just had a few more successes behind me, so I felt more confident about a temporary lapse in productivity—maybe I could force myself to do it. I do want to try it from the top down with a fresh novel, if I can ever finish THE DANDELION KNIGHT (and the graphic novel, which has to get finished first, and a couple of short stories whose deadlines might start to take precedence in the coming weeks, etc.), but I couldn’t take the thought of stemming the flow of words for that long. THE DANDELION KNIGHT is mostly written, now, in the very low-quality draft form that the snowflake method strives to avoid, and it will be completed and revised the hard way.

In the not-too-distant future, I’ll talk about Harry Stephen Keeler’s webwork plotting method. In the meantime, if you’re a writer who’s tried this method (and you’re not trying to sell it), testify in the comments.


One thought on “begin with the end in mind: some thoughts on the snowflake method

  1. Pingback:

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s