How We Really Write Novels

I’ve been reading a book titled Code Complete. It’s the kind of book that if you write any computer code at all, from a simple webpage to a large software project, you ought to read. For the rest of you, you would find if very dull. But as I was reading it I thought about the similarity between software development and writing a novel. There is a debate, if you will, among novelists between the plotters and the pantsers. The plotters say you ought to outline the book first, then begin writing. The pantsers just grab a few interesting characters and jump right into writing. Which is better?

One of the things Steve McConnell talks about in Code Complete is that no one ever develops software completely using a top down or a bottom up approach. The top down approach is similar to what plotters do, in that you start at a very high level of abstraction and subdivide it over and over until you reach the point where it is manageable chunks. The bottom up approach starts with the parts and you figure out how to tie them together into a useful product. In writing, the top down approach is given names, such as The Snowflake Method. But in writing, as with software development, no one does the whole thing as either top down or bottom up.

Realistically, a plotter may start out with the idea that he will write a mystery novel. That is about as abstract as you can get. He might follow the three act structure and develop scenes that fit it. But reaching that level of abstraction, it becomes much more difficult to continue refining from the top down. Likewise, a pantser might begin with a character and a problem, but after writing several pages she realizes where this book is headed. Her book is also a mystery and that means she has to be careful about what she reveals and she has to wait till the end to reveal why whoever did it did it. It also means that most of the book is going to be about someone searching for the answers. What happens is the plotter reaches a point he finds it easier to just write, so he can get to know his characters and the pantser reaches a point where she has an outline, whether she wrote it down on paper or not.

Another thing Steve McConnell mentioned is that we often talk about what we believe we should have done rather than what we actually did. The Snowflake Method, for example, talks about more and more refinement, all the way down to the final product. I’m sure Randy Ingermanson believes that’s what should be done. But in the real world, people don’t write that way. I just doesn’t work. Others talk about just writing about the characters and seeing where it takes them. I’m sure they believe that’s the way it ought to be, but I doubt anyone can do that for a whole book.

I’ve also been reading Richard Mabry’s Diagnosis Death. I’ve read his other books and one of the things that I’ve noticed about this novel is how similar it is to his other novels. They are all about a medical doctor who has someone out to get her, for some reason or another. And yet, Dr. Mabry has claimed to be a pantser. I don’t know him very well, but I know him well enough to believe he wouldn’t lie about that. Even so, he knows where his books as going.

So, my conclusion is that while we may believe plotting or writing by the seat of our pants is how it ought to be done, in reality, we do both. We will start a book with one method, then switch to another. Before we’re done, we may switch back and forth several times. There’s nothing wrong with that. In fact, that is probably the best thing we can do because it allows us to see the story from different perspectives, giving us a better book in the end.