To clarify how I work. By analogy.
The story as building site. An empty space.
There comes first an essential idea, a kind of story I want to tell, and it is very broad, ephemeral, but not quite. There are no blueprints. Maybe I know it is intended to be a house and not an office, a bungalow and not a skyscraper. But who knows?
First the ground is cleared and made ready, so I know the parameters of the space available, now I can build anything, but within these boundaries.
Then the foundation is poured, the support structure for the entire story. The essentials are placed. Then the bricks that will contain and support the space. First one brick, then another, then another … but again without drawings or plans, allowing the placing of the bricks to determine the location of the next and the next, until there is a foundational row on which to build.
In time, the bricklayer begins to see a pattern to the formation he has placed. Soon there is enough foundation to create a necessary picture of what must come next. In time, there comes a vision of the completed structure, what will appear at the end of it. To deny this would mean destroying everything that came before, tearing out all the previous work. When it is the previous work itself that makes the rest of it. If the early work is wrong, everything stemming from it will end up wrong.
When I begin a story, I have no plans and no scheme. Often all I have is a character who interests me, or maybe a single compelling scene. Possessed by Shadows began its life as one small scene – two men in a storm-buffeted tent on a mountain side with a dying woman between them; the questions who and why?
The current novel began as only the idea that it is only love, the only one of our human emotions, that can completely wreck a person, no matter how sane or rational or stable one is.
Those things constitute the cleared ground.
The making of the story is setting each word in each sentence so that it rests solidly on each word and each sentence preceding it. If I have made a mistake on the first page or pages, that mistake will become compounded over time, until the story falls apart from lack of proper support. I believe that writers who find themselves going back again and again revising entire manuscripts are forced to do so because a mistake in the early stages of construction become compounded and destructive because it wasn’t seen. (Neglecting the rebar weakens the wall, so a strong push later on knocks it down.)
I have always thought this obvious. Each sentence necessarily follows the one before, and each page the one before. If you make a mistake and make something yellow when it really would have been better red, then everything yellow after it will be wrong, and all of it thrown out to be replaced by red, as it ought to be have the first time.
For that reason, I do not leave a page behind until I am confident that it is exactly right. Which gives the next page a solid support from the start.
What I cannot know is what this steady progression of details may imply for the whole, because the next page is always a mystery, it is always empty until filled. Story-telling in this way is an accumulated effect, the revelation of the layers.
Returning to analogy. Each day the builder steps back and surveys what he has built so far. Each day the structure rises incrementally and defines itself. Ah hah! the builder says. It seemed to be a bungalow, but now I see it will make a better apartment building.
That is how the plot developed step by step to make a bigger story than I had first imagined.
Had I plunged ahead, placing bricks willy nilly, almost with literary whimsy, just to see what forms I could make, at some point I would be likely to see that I had made something possibly interesting, but probably insubstantial, a structure with no purpose.
Having to write the same story over and over means to me that it was built on unwitting mistakes.
So when I top out, or place the last stone, I am done and the reader can move in right away.
That is why I write the way I do. I am lazy and just want to do it once and get it right the first time.