Did you write Lexicon chronologically and later decide which order to publish the chapters in? Or did you write it in the order in which the chapters appear? Or something else?

Robert

I wrote scenes in the order they occurred to me and moved them around later. That wasn’t chronological order,
but wasn’t the published order, either, since I rewrite like a lunatic.

Actually, that’s misleading. I also write like a lunatic. It’s not just rewriting. When I had the basic idea
for Lexicon, the first thing I did was write 20,000 words that never ended up in the book.
I think a lot of writers do this. Many people have told me they started writing
a novel and loved it for ten or twenty thousand words and then lost interest. I always say it’s not you; it’s the
book. You haven’t lost steam because you’re not a good enough writer; it’s because your stupid story isn’t giving
you enough to work with. You had something good but went wrong and now you’re trying to decorate the
Sistene Chapel ceiling with crayons. Did Michelangelo use crayons? No. That ceiling would have sucked
if he’d had crayons. People would say, “That is one mediocre ceiling. I can’t believe I came all the way here to see it.”

So when that happened to me, I threw it all away except for a 500-word scene of a guy getting assaulted in
a bathroom and 1,000 words of a street hustler’s magic game gone wrong. I still found those interesting.

This time when I hit 20,000 words and began to hate it there was more to salvage: I had characters
like Eliot, Yeats, and Bronte, and a stronger idea of what people were doing and why. Everything else was
still terrible. I didn’t have a story so much as a bunch of different people doing different things.
But there was more there. So I cut back to 10k words and spent a couple of years writing and cutting and
rewriting my way up to 40k.

Then I threw it all out for the same 1,500 words I’d had before. I’d developed a lot more of the world,
but the whole thing still sucked for reasons I couldn’t identify. So I decided to try a new approach.

(During this time Machine Man went from idea to online serial to published book. It was nice to work on something
with a linear relationship between time spent writing and book length.)

This time, I made those simple two scenes the openings of Chapters 1 & 2 and spent the entire rest of each chapter
exploring them. With all the other stuff stripped away, it immediately felt more like a real story.
I ran it up to 80,000 words without too much trauma, relatively speaking,
sticking with this new format of alternating point-of-view chapters: Wil, Emily, Wil, Emily.

That became increasingly challenging as the ending loomed and I needed to bring story threads together.
For example, I would want to do something with a particular character in a particular time-frame at
a particular point in the story, but it wouldn’t be the right point-of-view chapter. I managed to make it
work anyway, kind of, and finished a first draft (110,000 words), but it was complicated and hard to follow,
with too much jumping around in time and space. It even had the worst kind of flashback, where first you
see something that doesn’t make any sense, then later the story is like, “Oh, so here’s what you needed to know
back then. It’s pretty great now, right?” No! It’s too late. You can’t retrospectively save a scene. I already
experienced it and felt bad.

The structure also made it impossible to change anything. My first drafts always need a lot of reworking in the back
half, since they evolve through a beautiful, natural, organic process of creative discovery, instead of from a plan
like a sane person would use. My structure was a Jenga tower of Babel where I couldn’t touch any part of it
without risking collapse the whole thing, because it was all interconnected and inflexible.

So I straightened out the timeline, moving scenes to where they made the most sense
from a story point of view, rather than the dictates of an alternating chapter structure.
That sounds neat and tidy, like you can click and drag a scene from one place to another and
it will snap into the right place, but the reality is more like operating on someone who has their big toe growing
out of their forehead. It’s messy, is what I’m saying. You create a lot of ragged edges. There may be some
crying involved.

Usual disclaimer: This process isn’t something I recommend. Ideally I would have an idea, plan the book,
and write out a first draft in chapter order. I’m just not smart enough to do that. That’s the only problem. I can’t guess in advance
what will be interesting about a story. I have to wade in there and figure it out from ground level. But
maybe you can!