« Done | Main | Ode to Librarians »

February 16, 2004

Novel Post-Mortem, Volume One

All right, now I've had some sleep. Allow me to share some thoughts about finishing up The Android's Dream:

* First, I'm amused (as always) at how the writing process works, at least in my case. I started writing The Android's Dream (henceforth to be acronymed as "TAD") at the beginning of May, 2003 and fully expected to be finished by the first of October, which is when the book was contractually due to the publisher; I had completed both Agent and Old Man's War in about three months, so the five-month time budget I gave myself for TAD seemed pretty generous.

But it turned out not to be, I think primarily because I didn't appreciate that I was writing a rather more complicated book than I'd tried writing before. Agent and OMW are both first-person, dialogue-driven and linear in their story-telling -- both novels are basically on a rail from start to finish and don't deviate from the course. Nothing wrong with that, of course (I like both of those novels a lot). But TAD is third-person, with multiple story lines, rather more description than I've ever attempted before, and a plot that has several layers (the story of the main character, a larger diplomatic struggle into which his actions are set, and so on and so forth). Like I said, complicated. The complication was added to by the fact I wrote without an outline -- a fact which I think is good (more on that later) -- but which also meant I spent a hell of a lot of time figuring out what came next.

End result: Nine-and-a-half months. If I had been carrying a baby, I would have been screaming to get the damn thing out -- and I sort of was. The last month of the writing was actually rather intense, in terms of output: Most of the back third of the book (about 35,000 words) in that time, over 23,000 words of of which were in the last five days, and 14,000 words in the last 20 hours or so. This after a couple of months where I was lucky to get a couple thousand words out.

The difference was that in the early months my brain was still working on how the plot would unfold, and in the last month the plot stopped unfolding and started folding back in on itself -- the "backside" I wrote about a couple of weeks ago, in which I as a writer know (more or less) everything that had to happen to get to the end of the book, so now it was a simple matter of writing it. And once I know exactly what I'm going to write, you know, stand back -- I write fast. It was gratifying, actually. After spending several months convincing myself that it was okay to write slowly because the story just needed to be written at the the pace it was going, it was a blast just to pound out that final third.

So writing this book was like eight months in first and second gear followed by a month in fifth gear and the final five days with a rocket engine strapped to my head. I don't know that I'd recommend this sort of writing technique to anyone else (I don't know that I'd recommend it to me), but in this case it's what worked. If you have the time and luxury to let a story write itself at the pace it wants to be written (and I did -- thank you, Patrick), you should do it, no matter the pace, and no matter the changes in speed along the way.

* Am I happy with it? I am indeed. Writingwise, I think it's pretty good, especially some of the final chapters, which are just jammed with action. And that's a good thing, since I sold the book as being action-packed. Structurally, I think it's sound as well, and I'm pleased to discover that I can write a third-person story with multiple plot lines. Now that I know I can, I do think (hope, pray) that subsequent excursions into the form will write themselves more quickly; this particular story-writing muscle got a workout. Pacingwise, I also think it's good -- it starts off with a bang (the six of you who attended my reading at Torcon 3 know about this), lets out some line to let the story develop, reels in the slack with a couple of action-oriented chapters, and then from there repeats the cycle of "little bit of slack, then lots of action" right up to the end.

I'm also pleased with the quality of the story. I don't like to contend that I am an entirely original thinker when it comes to plots; I enjoy taking basic plot ideas I know have worked before and try to put a new and interesting spin on them (OMW is an example of this: Your basic classic SF military tale with the twist being the age of the combatants). TAD has a lot of classic themes -- worlds in the balance, a hero on the run, people caught in the crossfire of power struggles above their heads -- but I there are a lot of story details and plot devices that haven't been used before, which will make it enjoyable to read and make people want to find out what comes next.

Is TAD about anything? Well, as I said over at By The Way, you could say it's about identity -- who we are as opposed to who people assume we are -- since several characters in the book have identity issues. But speaking as the author (yeah, yeah, what do I know?) I don't know that I would make a huge deal about the identity issues. The book is designed to be a fast, fun but not-entirely-stupid adventure story. Tonally speaking, if this book could be said to have a model, it would be the books of Carl Hiaasen or Gregory MacDonald -- mystery and thriller writers -- than any particular science fiction author. It's not meant to be a classic of literature, it's meant to be something you might read on a beach or an airplane. It's what I was aiming for.

I have no objection to writing books about ideas one day, if the opportunity/inclination comes about. But I promised Tor a book they could sell in supermarket racks, and also to the point, I wanted to write a book like that, too (I want to be huge! HUGE!). I think this book does that.

* As I mentioned, I typically write without an outline, and one of the things that means is that I'm often amazed at what I thought would be in the book that is not, and that some things that turn out to be integral to the book I hadn't even thought of prior to the writing. For an example, early on the book I imagined that there would be alien rebels who would help the hero in a "storm the castle" sort of action sequence; there are alien rebels, but they have very little to do directly with the hero, and they certainly don't help him storm any castles. Likewise, I imagined a rather significant section of the book would involve an information metaphor that large-scale computer systems use to communicate with each other, but I never got around to using it because it didn't make sense to shoehorn it into the story. Not every cool idea (or accurately, every idea you think is cool) is going to work.

On the flip side of this, there was one small character detail which turned out to be hugely useful with the construction of the book; I returned to it again and again and it fact it became one of the spines upon which the story is supported. It's significant enough that I wish I could say I planned it. Once the book comes out I'll let you guys try to figure out which detail it might be.

As an aside to this, there are a number of places where I think it could have been interesting to delve deeply into one detail or another, but I chose not to as a writer (one of these involves a specific character, who exists in a certain state). I didn't go into these bits generally for two reasons: One, there's a small chance that I'll create other books in this universe, and I'd like to leave some avenues open for later exploration. Two, I think it's fun to have a universe that's incomplete -- not sketchy or underwritten, but rather where not every single thing is explained in detail. It gives the reader things to think about in an undirected sense. That is to say, you as the author aren't pointing to something going "Look! This is something to ponder!" Rather, readers find something and go "I wonder why that is," and let their imaginations run with it. I mean, hopefully. They might also go "Damn it, why doesn't he explain this?!?" But that's a risk I'm willing to take.

* When does it come out? Well, presuming Tor likes it, my understanding is that it will hit the shelves sometime around November 2005. Please understand this is an extremely tentative prediction on my part; it's up to Tor in its bibliographical wisdom to place it on the schedule.

Does this mean you'll have to wait at least 18 months to read it? Well, yeah, basically. I don't expect I'll be placing it online in any capacity between now and then. But it's not as if you won't have enough other stuff from me between now and then. In fact, I'm likely to have some announcements coming on that subject some time in the near future. But for now, good things (or at least things I write) come to those who wait.

Posted by john at February 16, 2004 04:31 PM

Trackback Pings

TrackBack URL for this entry:

Listed below are links to weblogs that reference Novel Post-Mortem, Volume One:

» Science Fiction Double Feature from Weblogging Considered Harmful
A couple of webloggers that I follow, Peter Milan and John Scalzi, have both just finished writing novels over this... [Read More]

Tracked on February 18, 2004 09:40 AM

» no rewrite? from postscript
A few weeks ago, a friend of mine finished the first draft of her first novel. “Yay,” I said, expecting to read it and give notes for the inevitable first rewrite. But no. Instead, after a cursory run-through for line... [Read More]

Tracked on February 20, 2004 12:35 AM