How do you write a novel, again?
Oct. 19th, 2003 09:35 pm I have been thinking about the ways I need to change my writing process with this new novel. I wrote two novels using one particular process, but I'm using a different method this time, and I'm still sorting out how my processes must change to reflect this new method.
Specifically: I wrote my last two novels straight from the beginning to the end. I used WordPerfect, and stored the novels in files, one file for each chapter. I used the WordPerfect master document process. When I wanted to look at the novel as a whole, I would open the master document, which is simply a string of macros that tell the program what files to put together to create the novel. When you generate the master document, it looks at the list ("Stick the prologue here, stick chapter 1 here, stick chapter 2 here . . . ") and plugs the files in. Hey presto, instant novel. But you wouldn't make any changes to the novel as a whole, once you generated it. All changes were made instead to the individual chapter files.
Each chapter file had a corresponding dump file, which was sort of a shadow file: in it I put all the false starts. If i was working on a paragraph, and it took three attempts to craft it correctly, I'd stick the first two attempts in the dump file. These files were called "dump01" "dump02," etc.
Now as for the method I used for research: I have a very strongly eidetic memory. I read lots of books when researching my novels, esp. Wild Swans, but I don't think I took a single page of notes. I have, upon occasion, astonished Kij by remembering a phrase from a book I read ten years previously, and I can go to the book on the shelf and quickly page right to the paragraph I want. My practice was to write with my research books right on the desk beside me or on my lap. When I was writing about Patience's cottage and knew I wanted to describe what hardware would be hanging in her fireplace, I would remember that the passage I wanted was in Colonial Life and Times, and I would be able to page to it pretty quickly. For Swans in particular, what this meant was that I had two sets of research books, 17th century and 20th century, and every two weeks, as I would finish one chapter, I would clear all the books from one era off my desk, and switch to the other set of research books.
Oh wait, I take that back: With Swans I did do an elaborate calendar, recording events from several historical sources (And the Band Played On and The Gay Decades). This was a tip I got from Tim Powers: put historical information you're trying to work into the story in pen, and your fictional events, which you can re-arrange, in pencil.
The process I'm using to write this new ice palace book is totally different. And as a result, I'm re-examining all the processes I've used in the past, and thinking about whether I want to try something I've never tried before. Specifically, I'm writing bits and pieces all over the book, like piecing together scraps to make quilt blocks, with the intention of stitching these quilt blocks together. That old system I used just doesn't work anymore. I'm not working in chapters, and I'm not working in order, so I can't keep track of stuff by simply numbering it. What to do?
kijjohnson and I had our weekly call today, and I pumped her at great length about her process, since this quilting method, after all, was her suggestion. She keeps three files: the novel itself, which is all one big file, a research file, and a diary file. The diary file, as she describes it, seems to be sort of a meta-musing about the book as it progresses, like the entries I've been making about writing this book on my livejournal, that I've been printing out and keeping. The entries are in chronological order, and include her musings about theme, reminders to herself that she needs a certain scene to accomplish X, suggestions that she trace a story arc in this way, etc.
The novel itself is kept all in one big file. She assembles her fragments in what she thinks is the rough order they'll be in, and she tries to see connections. Then she identifies what she needs to write to connect the fragments.
She also has a large research file, which might be 100+ pages by the time the book is done. As she reads her research books, she flags anything she wants with colored flags. She goes back and writes all those notes into her research file, sorting them into rough categories (clothing, kitchen, food, travel, weapons, etc.). Each note has at the end a code indicating what book she culled the note from, and the page number.
She prints out her novel fragments and writes all over them, another thing I have never done. I print out a first draft, but I do all my revisions on screen only, which is probably why graduate students who study my work someday (if there are any) will hate me, because this rather obscures my editing process, for historical purposes.
She also puts plot fragments on post-it notes and experiments with rearranging them. Some are cause and effect, and those are easy to order, but others it's not so clear where they go in, and the post it notes let her experiment with a variety of arrangements. She talked about sometimes color coding stuff with highlighters, too. Not sure if that's in her novel draft or her diary file. For the Fox book for example, she had a different highlighter color for each of her three main viewpoint characters.
I am thinking I need to install MS Word and start using that instead of WordPerfect. Yeah, yeah, I like WordPerfect better, but 1) we're using Word at the office now, so it's the program I'm using every day 2) it's the default for everybody, so editors will expect me to submit the book in Word and 3) I need to get Word for transacting HPEF board business. I am thinking about trying some of these other tools that Kij mentioned. Would I really need to write all my research down? I may be depending more on live interviews this book than I ever have before (i.e., talking with ice palace architects, structural engineers, etc.) Would I find writing the research down helpful, or an unnecessary time waster because of my eidetic memory? Not sure. Will have to think about that. The diary file sounds intriguing, too; it's essentially what I've been doing with these LiveJournal entries. Could continue to do that, and just copy what I want to make public to LJ.
Peg
Specifically: I wrote my last two novels straight from the beginning to the end. I used WordPerfect, and stored the novels in files, one file for each chapter. I used the WordPerfect master document process. When I wanted to look at the novel as a whole, I would open the master document, which is simply a string of macros that tell the program what files to put together to create the novel. When you generate the master document, it looks at the list ("Stick the prologue here, stick chapter 1 here, stick chapter 2 here . . . ") and plugs the files in. Hey presto, instant novel. But you wouldn't make any changes to the novel as a whole, once you generated it. All changes were made instead to the individual chapter files.
Each chapter file had a corresponding dump file, which was sort of a shadow file: in it I put all the false starts. If i was working on a paragraph, and it took three attempts to craft it correctly, I'd stick the first two attempts in the dump file. These files were called "dump01" "dump02," etc.
Now as for the method I used for research: I have a very strongly eidetic memory. I read lots of books when researching my novels, esp. Wild Swans, but I don't think I took a single page of notes. I have, upon occasion, astonished Kij by remembering a phrase from a book I read ten years previously, and I can go to the book on the shelf and quickly page right to the paragraph I want. My practice was to write with my research books right on the desk beside me or on my lap. When I was writing about Patience's cottage and knew I wanted to describe what hardware would be hanging in her fireplace, I would remember that the passage I wanted was in Colonial Life and Times, and I would be able to page to it pretty quickly. For Swans in particular, what this meant was that I had two sets of research books, 17th century and 20th century, and every two weeks, as I would finish one chapter, I would clear all the books from one era off my desk, and switch to the other set of research books.
Oh wait, I take that back: With Swans I did do an elaborate calendar, recording events from several historical sources (And the Band Played On and The Gay Decades). This was a tip I got from Tim Powers: put historical information you're trying to work into the story in pen, and your fictional events, which you can re-arrange, in pencil.
The process I'm using to write this new ice palace book is totally different. And as a result, I'm re-examining all the processes I've used in the past, and thinking about whether I want to try something I've never tried before. Specifically, I'm writing bits and pieces all over the book, like piecing together scraps to make quilt blocks, with the intention of stitching these quilt blocks together. That old system I used just doesn't work anymore. I'm not working in chapters, and I'm not working in order, so I can't keep track of stuff by simply numbering it. What to do?
The novel itself is kept all in one big file. She assembles her fragments in what she thinks is the rough order they'll be in, and she tries to see connections. Then she identifies what she needs to write to connect the fragments.
She also has a large research file, which might be 100+ pages by the time the book is done. As she reads her research books, she flags anything she wants with colored flags. She goes back and writes all those notes into her research file, sorting them into rough categories (clothing, kitchen, food, travel, weapons, etc.). Each note has at the end a code indicating what book she culled the note from, and the page number.
She prints out her novel fragments and writes all over them, another thing I have never done. I print out a first draft, but I do all my revisions on screen only, which is probably why graduate students who study my work someday (if there are any) will hate me, because this rather obscures my editing process, for historical purposes.
She also puts plot fragments on post-it notes and experiments with rearranging them. Some are cause and effect, and those are easy to order, but others it's not so clear where they go in, and the post it notes let her experiment with a variety of arrangements. She talked about sometimes color coding stuff with highlighters, too. Not sure if that's in her novel draft or her diary file. For the Fox book for example, she had a different highlighter color for each of her three main viewpoint characters.
I am thinking I need to install MS Word and start using that instead of WordPerfect. Yeah, yeah, I like WordPerfect better, but 1) we're using Word at the office now, so it's the program I'm using every day 2) it's the default for everybody, so editors will expect me to submit the book in Word and 3) I need to get Word for transacting HPEF board business. I am thinking about trying some of these other tools that Kij mentioned. Would I really need to write all my research down? I may be depending more on live interviews this book than I ever have before (i.e., talking with ice palace architects, structural engineers, etc.) Would I find writing the research down helpful, or an unnecessary time waster because of my eidetic memory? Not sure. Will have to think about that. The diary file sounds intriguing, too; it's essentially what I've been doing with these LiveJournal entries. Could continue to do that, and just copy what I want to make public to LJ.
Peg
(no subject)
Date: 2003-10-19 07:50 pm (UTC)(no subject)
Date: 2003-10-19 07:57 pm (UTC)Every one of the damned things is different for me: I've never had two novels work the same way. Some are linear, some are crazy quilt. Some have hardcore outlines, and some just--come.
I take comfort in the Gene Wolfe quote that runs, "You never learn how to write a novel. You learn how to write this novel."
So this one being a problem child is probably as much a sign of growth as anything, and all-over positive. I know, comforting, right?
If you ever wanna geek over outlining methods, I'd be happy to show you what I've arrived at--the vanishing outline, I call it--where the outline (if something this fluid can be called an outline) is consumed as the book progresses, but it allows me to jump ahead when I have crazy-quilt inspiration.
I don't save separate files unless I cut a big chunk, but what I do is save a new dated file daily, so I can find the old stuff if I need it. Including consumed outline bits if I need them back.
I know I'm teaching my grandmother to suck eggs and all that, but if there's anything I can do to help, let me know. For me, there's nothing more traumatic than a story that's trying to break the mold.
And I find watching your writing process very valuable, personally. And also a little bit comforting. *g*
(no subject)
Date: 2003-10-20 06:33 am (UTC)Thanks for the encouragement.
(no subject)
Date: 2003-10-20 07:19 am (UTC)(no subject)
Date: 2003-10-20 09:06 am (UTC)--I suspect it's one more reason why writing can be learned, but not really taught.
*\o/* we have cheerleaders for all seasons *g*
(no subject)
Date: 2003-10-19 08:38 pm (UTC)Oddly enough, I used only the vaguest outline for the linear writing, but had a nicely detailed outline for the crazy quilt. I came up with my CQ outline after a couple of early chapters were mostly completed and I had enough quilt bits that I'd figured out what was going to happen in the plot. The outline helped me stick pieces where I needed them, but the story determined the outline as much as the outline determined the story, which was a new thing for me.
Pros: Linear writing worked well because I was retelling a fairy-tale and I already knew where I was going next. Plus there was the whole "finished chapter 4, on to chapter 5!" rush, which is pleasurable and makes my inner bean-counter happy -- thus far finished, thus far to go, see the progress. The crazy quilt approach, though, worked better creatively, and was probably due to the fact I wasn't intending to write a fanfic -- I just kept assembling these little scenes and realized I had a big story. It was more of a rush, because I'd drop in and out of the story wherever I felt the urge, but when I wasn't feeling particularly urge-y I had Something To Do -- the filler work to connect the pieces I'd written before.
Cons: With the linear writing, I made myself plow through the parts I wasn't enjoying as much, but with the crazy quilt approach, I put off scenes I didn't want to write -- and was stuck with writing them, all in one unsavory lump, at the end! However, with the linear writing, the rewrite process is much more rocky. Hopping back and forth in the CQ approach meant I was constantly monitoring the whole book, where the linear approach meant that I had to keep going forward, and not tinker with what I'd done until I'd completed the first draft. I still have mucho rewriting to do on that one.
Practical matters: The crazy quilt had a file for each chapter and another file for "extra" bits that didn't get used. The linear was all in one file, and just saved with a new date in the filename every day.
A side note: I personally have a great memory for things-in-books but a terrible memory for what I hear, so notes for interviews, etc. are quite necessary for me. It doesn't have to be very organized, I just have to write it down, and then it's a written thing and sticks in my head. But your memory could be better than mine. :)
(no subject)
Date: 2003-10-19 08:43 pm (UTC)These days MS Word can easily handle book-length documents.
I essentially write in the patchwork way you describe. I am constantly moving sections around, chapters around...even clever phrases around. Almost all the time everything goes somewhere, though. A section might move, but when I write it I know where in the book it goes. There are exceptions, and I create a final chapter of my book called "Orphanage" where I put the homeless paragraphs, sections, one liners, clever ideas, and the like.
My guess is that my process will be harder to make work in fiction, where there is a much stronger narrative structure. But it does sound somewhat what Kij does.
B
(no subject)
Date: 2003-10-19 10:12 pm (UTC)(no subject)
Date: 2003-10-19 10:26 pm (UTC)Maria
(no subject)
Date: 2003-10-20 01:52 am (UTC)(no subject)
Date: 2003-10-20 06:25 am (UTC)I am mostly amazed at how organized you and Kij are. I suppose I am organized in some ways, but in others I am pretty haphazard--perhaps because I'm in the draft stage and not so concerned with continuity yet. My research is only loosely organized, depending on which notebook has room for another Xeroxed chapter.
Before I started my "real" first draft, I wrote a series of fragments, but I did not intend to actually use most of them in the draft. A couple of them made it in, and the rest were just me getting used to the characters and how I wanted to write them.
(no subject)
Date: 2003-10-20 07:27 am (UTC)I'm using the full-power ahead draft method right now. When I get stuck, I go back, as little as possible, and rewrite from there, saving bits I discard in another file if I think they're good. I didn't outline, except in my head. I did have a vague shape in mind, but the pacing is way off from what it needs to be in the rewrite. I figured out a lot of things in the draft, and it took a lot of wordage to figure them out, so nothing much happens but character interaction for large chunks.
Rewrite, I'm going to have to come up with another method. I've never done a full scale novel rewrite before, so it will be a new adventure.
(no subject)
Date: 2003-10-20 03:19 pm (UTC)Regardless, thank you much for sharing these posts. I admire your courage and your choice to post some very personal things.
(no subject)
Date: 2003-10-20 07:28 pm (UTC)Cheers,
Peg
(no subject)
Date: 2003-10-20 06:45 pm (UTC)I've never successfully finished a book, but quilting I understand. =)
(no subject)
Date: 2003-10-20 08:38 pm (UTC)I've never written a novel, but when I was writing my 76-page fic (which, for me, was darn near to a novel; by far the longest thing I'd ever done), I did what Kij did. I wrote in one giant file (I didn't even mark the chapters until everything was done and I had a better feel for where the breaks really needed to be) and wrote scenes out of order. I kept a separate file with notes and research. The *general* progression was chronological, but there were times when I just couldn't do the 'next' section, or where a later bit just clamored to be written first. However, I'm usually much more like you - I vastly prefer to write from beginning to end, so this was something new for me.
Usually when I write, I jot down a general plot overview on a scrap piece of paper, and reference that as I go along, so I can walk away and still remember what my idea was when I come back. Because this was much longer, I had a basic outline, which then blossomed into several notebook pages of tiny scribble as I used the "plot outline" concept for most main scenes. I'd scribble/brainstorm the next scene I wanted to write, and then use that to actually *type* it out into the computer.
I don't know what will work best for you, of course, but that's what I did on my much-smaller-than-your-novel work.
BTW, if you decide to use LJ as a tool, you could start a Memories category for your novel notes, and that way you could find the posts more easily when you wanted to reference them later.
Quilting and writing
Date: 2003-10-21 09:09 am (UTC)As to writing, I wrote government documents on computer security about 10 years ago using WordPerfect, the standard at the time. How the document was organized depended on if I was working alone or working with a group (and ever changing group). I think I outlined either way, but the group effort meant a lot more shifting around of chapters or dumping things altogether. I always printing out the document in progress for comments and corrections.
Before I was laid off, I was a software tester and wrote test plans and user documents. The test plans relied heavily on culling from documents produced by the system analyst and programmers, when they (the programmers) felt like it. I suppose it was a bit more like editing, as I was the one who did the sanity check and asked, "Are you aware that these two documents don't agree on what's being done?" Which happened more often than one would expect, since the system analyst talked to the customer and the programmers talked to themselves. My favorite was the programmer saying, "No problem" to however the system analyst designed the section to work and then handing me something different saying it won't work that way. Major revisions there, but I ramble.
Doing the user's document was usually following an outline as I needed to cover all the new functionality as well as how to trouble shoot. I would start it when I got a new project and put notes in as to what should be included, since a project could drag on for months. By that time I was so familiar with the functionality, I needed remember that the user probably wasn't.
Thanks
Date: 2003-10-25 04:58 pm (UTC)You are very generous (not to mention brave) to put this account of you creative process where everyone can see and learn from them. It is nice to have a solid example of just how much the process varies from person to person.
Thanks again!