Graphic by Peg Ihinger
When I’m between books, I frequently check out different writing systems, hoping I’ll find something that will make starting up the next book easier. Lately, I’ve run across quite a few “how to write a novel” systems that take a fractal approach to writing a novel.
For those unfamiliar with the term, fractals are geometric patterns that repeat at different scales. Like the Sierpinski triangle, https://en.wikipedia.org/wiki/Sierpi%C5%84ski_triangle
At first glance, this approach seems well-suited to writing a novel. All the fractal writing systems I’ve seen start with either a one-sentence summary of the plot (most of them) or a one-sentence introduction of the main character (a few). The system then recommends some method of expanding that sentence into a few more, ranging from a short paragraph of plot description or character background, to specific structures like a three-act or hero’s journey variant. That summary gets expanded again (usually to a one- or two-page outline, but sometimes to a set word count), and then again into a longer outline, rinse and repeat until the writer reaches the point of actually writing the book. Some methods stop expanding once the outline/character description reaches a set number of pages (often five, ten, or twenty-five). Others get to a “long form” plot outline and then keep going, breaking that down into chapter outlines, and then into scenes.
Interestingly, there is always some point in the process where the system recommends stopping work on the plot (or the character, if that was the focus) and switching gears to the other one—”Now stop and make up some characters/ plot based pm what you’ve already made up.” Then they go back to expanding the summary to the next level down.
In theory, this results in a really detailed outline that the writer can work from with great confidence. In practice, I find that, like all other writing advice, it works really well for some writers and really poorly for others. And for some, like me, it only works sideways, rather than as intended.
I have always felt that this kind of growing-plot-from-a-sentence ought to work for me, especially since I often start with a seed-crystal idea and grow it into a five-to-ten-page outline plus notes. How I grow it, however, is not as neat and tidy as any of the systems. More important, all of the fractal-writing-system methods I’ve encountered so far take a top-down approach. You start with that one-sentence summary of the idea for the book.
I am a natural novelist. I have terrible, horrible problems summing up anything in just one page, let alone just one sentence. Especially when all I have is a tiny seed-crystal of an idea, because that seed-crystal is never a plot summary or a character introduction. It’s a seed. It doesn’t need more and more detailed versions of itself. It needs to grow roots and stems and leaves and branches that don’t anything like the little round thing that I started with.
What I can do is write a sentence that might go somewhere in the book: “Mother taught me to be polite to dragons.” “‘My lady, we have a problem.’ ‘Just one?'” “The gentleman in the silk top hat and fine wool coat looked startlingly out of place in the low-class tavern.” “People are real nice to a double-seventh son.”
None of the fractal writing systems I’ve seen ever talk about expanding that sentence into a snippet of a scene, and then into a chapter summary, and so on, until you get to the whole thing. (I keep saying “that I’ve seen” because the internet is very large, and I haven’t seen anything like all of it.)
Both fractals and stories can build in any direction—bottom up, top down, or middle out. And while I don’t think it works (certainly not for me) to whittle the complexity of fractal patterns down to a rigid system of repeated expansion, fractal development does work for me if I approach it more loosely.
Even the systems recognize that developing a story takes more than repeatedly expanding a log-line until you’ve gone from one concise summary sentence to 400 pages of detailed description, incident, dialog, etc. That’s what the “Now stop and switch gears” moment is. Development has to switch focus periodically, because expanding and developing only the plot (or only the characters) isn’t enough, no matter which direction you’re working in.
It’s that switching of focus that lets me go from a random seed-crystal sentence to, eventually, a finished manuscript (all 400-plus pages, submission outlines, and so on). If I look at the neat systematic recommendations (do precisely this in exactly that order), I get nowhere. But I can consider a gentleman sitting in a scuzzy tavern, and figure out that I need to go research Regency England right now. And then I find a street map of London in the early 1800s, a copy of The 1811 Dictionary of the Vulgar Tongue, and a thin book on caravans (the equivalent of a modern-day camper or RV in Regency through Victorian England), and I get an idea of who’s watching the guy and why.
Bouncing back and forth, from big picture background (Regency England shortly after Waterloo) to character background (my gentleman was an upper-class Brit who worked as a spy during the war; he’s being watched by a street urchin) to plot details (which can be from the middle, beginning, or end of the story) eventually creates patterns, and from there, it’s a matter of filling in empty spaces. It’s still kind of fractal, even if it’s not particularly systematic.
The main thing is to keep inventing bits that fit the pattern that’s developing. It doesn’t matter whether they’re character bits, plot bits, background bits, snippets of dialog or description or action or…anything. What matters is accumulating enough bits to recognize the pattern.
The examples of seed-crystals you give look familiar. They’re like mine, except that mine are more abstract rather than being bits of narration or dialog.
Only they’re not seed-crystals, or at least not for me. They’re not crystalline. Nor are they seed-like, with proto-leaves and a proto-root encapsulated within. They’re more like a bit of the stuff in the solution, and need a separate seed-crystal.
I keep tripping over this unspoken assumption that it’s somehow easy-peasy to take a bit like that and turn it into a seed-crystal, or to treat it like a leaf that can be rooted and turned into a whole plant, the way some plants can be vegetatively propagated from a leaf, rather than from seeds.
Only it isn’t. Not for me. I need a seed-crystal that’s actually crystalline, in addition to the Cool Bit. And coming up with those isn’t trivial.
This looks like an interesting approach, though it’s no more for me than, evidently, for Deep.
I often start with a theme, and I don’t really see how this approach would work with that. I start the rest of the time with something that feels like there’s a story in it, and I just start writing. As long as I have enough in my head to go on…I go on. To the end, eventually. If at some point I don’t have enough, I put it aside, hoping to try again some other time.
(Since having enough to work on has never been a problem for me – until last year’s frights and traumas, anyway – I’ve never worried about my personal “abandonware.”)
I am unfortunately in the class of writers where, if I write a detailed short-form of the story, I have probably killed my motivation to write the long form. I think this is fairly common, and such people should stay far, far away from fractal systems.
_Zookeeper_ started with me looking down into an enclosure at the National Zoo and thinking, if I had to keep alien animals in there, how would I do it? I can readily imagine my protagonist doing the same thing–though she doesn’t, because it quickly became apparent that the enclosure has to be completely sealed off and viewed through windows. So the seed-scene isn’t even in the novel.
I went home and wrote a couple pages, at which point I had a protagonist. Sort of. I didn’t find out she was a telepath for several chapters (in my defense, she never thinks about this–she’s in denial).
A lot of needed setting work got done while…writing book 2, so I have to go back and redo all the space scenes in book 1. Oops. Perhaps it’s lucky it wasn’t published or I’d be stuck with anti-gravity and a gnawing feeling that if you can do *that*, space combat looks nothing like what we imagine it does.
A previous fantasy novel came about because I was cleaning up papers and found two pages, clearly in my handwriting, showing a highly charged political meeting among wicked magicians. They must have been at least ten years old and I didn’t remember writing them…. I typed them in and ran from there. Somewhat later I discarded that whole scene, so it’s not in the novel either–the novel’s in tight third and my protag isn’t present for that meeting, also the “reader doesn’t know who these people are” gimmick wasn’t working for me.
That one has a setting in which I’d run roleplaying games, though, so a lot of worldbuilding was already in place. I had a finished first draft in four months and I think the pre-worldbuilding is why.
Dunno if these are seeds or crystals or what. There are a *lot* of those two-page bits on the computer and in the papers bin, and most went nowhere. A few took 4-5 chapters to go nowhere: both of my NaNoWriMo attempts are in that class. Something distinguishes the ones that became novels. I have now typed and deleted three theories about that; honestly I don’t know.
I am also the sort of writer where writing down any sort of detailed outline kills the story for me. I can have a vague idea of where I’m going generally and a specific concept of where I’m going next, but if I have any more detail than that the story dies. If I have any less detail than that (such as not knowing what comes next even though I know where I’m going), the story also dies. It’s a challenging balance to manage.
Your process—snippets, sketches, pingpong, and gap-filling—sounds a lot like what I do.
Me too.
In fact, filling gaps with overarching ideas, when I get them, is the hardest.
I tend to think of it in terms of puzzle pieces. I often start with a concept for a character or situation that feels interesting, then add connecting bits (who is this character, exactly? what’s the setting? who else is involved?) until I have enough to start writing the story. Sometimes a concept floats around in my head for a long time before I find the right pieces to connect to it and make it come to life.
I realized in thinking more about this that if I feel I have *invented* an ending I’m likely to think poorly of it, which is demotivating. I need to *discover* an ending, and that requires writing a good hunk of the story so it has somewhere to stand.
I did not know what the ending of _Zookeeper_ was until the next to last chapter, which took a shocking (to me, anyway) turn. But I had no doubt whatsoever that those events happened: and once I wrote them, I could see what the ending necessarily had to be.
Currently poking at a couple scenes from Book 3 and I have *no idea* where it’s going. I only wish I knew what distinguishes the four stories where that process produced an ending from the pile of others where it did not. I’d feel more confident.
The only thing I think I’ve learned is that I mustn’t write past that uneasy “I need to know more about X” feeling or I may end up with something like my first NaNoWriMo where there IS no answer to “what is my protagonist’s backstory?” At least, I have never been able to find an answer with any ring of conviction, and while I like parts of this quite a bit, it seems unfixable.
I send the hero to school, and the story unfolds more neatly than when I know the overarching plot of a niece discovering the point of what her aunt did, and helping bring it to fruition.
Like Gaiman, Asimov, and Tolkien, I write to see what happens next. If I have an outline or any other means of foreseeing the ending, I’ve any motivation to continue.
… I’ve *lost* any motivation to continue.
These posts really should be editable.
Yes, me too.
It’s apparent now that book 3 will be about putting eight students from disparate background into a closed environment and putting them under a whole lot of stress. Surely, one of them will be some kind of trouble. Very likely more than one. But until I know them all–currently only three–I don’t know what kind of trouble, and I can’t invent an answer in advance or it will forever feel like I forced it.
I guess, thematically, they should get into human kinds of trouble, and not the same kinds we’ve seen in human/alien interactions. I am “allowed” to think about things like that. But if I said right now, there’s a student who …. man, I am not going there even in a blog comment.
For me, I can know the end I’m writing toward, generally speaking–but if I know too much about how to get to the ending, or too little about what is immediately ahead of me in the story, I lose the motivation to write.
I’ve found the big reasons outlines are killers for me is that sequencing doesn’t work for me and I do not care about external plots, just the character ones.
Which means I care about cause and effect and pingpong around character big moments and their causes and effects until a critical mass of framework emerges. Framework is distinct from structure and sequence because it has zilch to do with how the details get presented and everything to do with my authorial understanding of what happened and what it means from the perspective of everybody involved.