Stylistic Annealing
Why Even Bad Rules Lead to Better Results
Don't use adverbs. Don't use adjectives. Don't use "was" or "like" or "which".
Don't use filter verbs. Don't use intensifiers. Don't use complex, polysyllabic, latinate words.
Don't be vague, be specific. Don't use abstractions, use concrete images. Don't be repetitive, redundant or formulaic, vary your style.
Don't follow these rules all the time — first, do not harm.
The pendulum swings as you ascend the mountain. Before you embark, you don't even realize there are rules, let alone what they are. Then your eyes are opened, and with each step you follow every prescript with naïve dilligence. Then, at some point, you realize the rules are kinda bullshit — guidelines at best, a prison of elitist convention at worst — and perhaps you pride yourself on this insight and say once you know the rules, you can break them. We all know this story.
I think there's a sequel — eventually your sense of this game called writing becomes so refined that you discard any need to define your approach with respect to those rules. You aren't following nor breaking them.
For a long time now, I've found this talk of breaking the rules uncompelling, without knowing quite why. But there's no glorious rebellion in breaking rules that have no power over you, is there?
To my eye, what most call "breaking the rules" is in fact just adhering to truer, more fundamental principles. The ostensible rules were wrought by hands that only partly understood them. Perhaps the procession of the perihelion of Mercury breaks Newton's gravity, but that's because relativity is the deeper law.
I'm waffling. I want to talk about the process of writing better prose. Again. By form, an essay of this type inevitably begins by insisting that they aren't telling you what to do, perish the thought, and any suggestion herein of how you "should" write ought be disregarded if you feel it makes for a better story.
But what if we let the pendulum swing back to the other extreme? What if we say no actually, you should follow the rules, even if it's hard, even if you just want to do what comes easier, even if it seems pointless?
The "it depends" enlightenment leads a certain fallacy of nuance. With nearly all of the petty rules I listed off at the start, you can at least see where they're coming from. Are adverbs always inherently bad? No. Is the sentence that avoids adverbs always better than the one that doesn't? No. Yet nonetheless, does there exist a certain weakness of expression shored up by purging adverbs? I contend yes, and for that I can appeal to unerring popularity and authority.
"Adverbs good" is wrong, "adverbs bad" is wrong, so clearly the truth lies in the middle, right? Simply recognize where adverbs are weak, and only cut them if it makes things better.
What flaws the above reasoning is a misunderstanding of the process as a dichotomy, of viewing what happens as simple replacement of sentence A with sentence B, and thinking the problem is determining when A < B and vice versa.
You're not giving yourself enough credit.
Negatives are distinctly uninstructive. If you can't use "was", what do you use instead? Applying these rules with more than the chainsaw-like grace of a find/replace will take work. If you write a sentence and find it breaks the rules, it is only one sentence. But if you seek alternatives, the list of possible revisions is literally limited only by your imagination.
If someone says blindly following rules makes your writing worse, say skill issue.
Imagine yourself wandering a forest when it begins to rain heavily. The ground beneath you floods. To escape drowning, you climb a tree. A wise man floats by and smugly says, "Don't climb trees."
The reasons why are easily imagined. Branches can snap under the weight of a person, especially if the winds are particularly strong. If you're in a deep valley and the tree's not tall enough, it might not even let you escape the water. And if all else succeeds, you're still going to get drenched.
If you follow his advice, do you just hop down and lay in the mud while the waters rise? No, you'd look for somewhere else to go. You'd climb to safety atop a hill, you'd scan for shelter, you'd find the best alternative.
When you follow stringent rules, you always lose something; it might even be important.
But the value of the rules isn't that they make writing better, it's that they make writing worse. Unless you compensate. Then they're training weights.
So alright. Obviously, it's hard to write well while following bad rules. But how could it be any better than simply writing well, without all the extra steps?
Think about it in statistical terms. There would be no prescription if what's forbidden didn't happen naturally and often. A sentence laden with vices the rules decry is so often the first thing that comes to you. Measure that quick attempt against the list of all the other possibilities. Even if the sentence gets the job done, even if it's good, do you believe there isn't a better alternative out there, somewhere?
Rules knock you out of local optima, and force you to search farther afield. They prompt you to exercise your skills. As they say, what is written without effort is read without pleasure.
Annealing is a process of heating a material till its structure breaks down, endowing atoms with energy enough to migrate to new positions; then as it cools, it recrystalizes into a new, improved form.
Rules are a source of friction, and friction is a source of heat. If my defense of the rules amounts to no more than "the real power was inside you all along", why use the rules at all? Why not just anneal the words directly?
Open a blank page and write something. No prompt, no inspiration. It can be anything. After all, famously, what makes anything good is the execution, not the idea, so why constrain yourself with external ideas? I'm gesturing at the obvious. There's a reason blank pages feature in so many writer in-jokes.
After such a provocative defense of bad rules, allow me to undercut it all with a concession that actually, the rules themselves do, in fact, matter. If you can use a bad rule for a helpful nudge, you can use a good rule to fly.
Rules are prompts, and they prompt you toward something. A starting point for revision, the place to look first when a passage is underwrought and you just don't know why.
Recall where this essay began. If the rules are approximations of deeper underlying laws, well, what would be a more refined approximation? What general model spits out "don't use 'like' or 'as'" as a special case?
Rules are roads and they all lead to the archetypal form.
You'll find each of the rules I listed off at the start synergize together. They all reflect an underlying tendency: an adoration for strong verbs and specific images, coupled with a disdain for anything that filters or frays that focus.
What's the most general advice on prose I can give? Go out and read. Find the book with the best prose you've ever seen, and comb through until you find your favorite line. Search until you find a perfect sentence.
And then write it. Over and over. Don't repeat its contents, figure out its form, the arrangement of parts that would be preserved if you played madlibs with the word choice, the way a parody song retains the flow of the original.
Write, and strive with every sentence to replicate the archetypal form of that perfect sentence.
I'm being hyperbolic. You'll fail at this, and remember that last rule? First, do no harm. What you'll soon discover is that the perfect sentence sounds good once, but twice in a row? Ten times? The perfect sentence was great at conveying one thing, but is totally ill-suited when wrapped around a different meaning. If your favorite line is something particularly simple and understated, then its true power comes from all the setup that came before. In order to give a scene the perfect final line, you need to first write the perfect scene.
True perfection isn't rigid, it's fluid. Do you actually have a favorite line, or a long list of quotes you treasure? Lay them out, and you'll find common themes. Perhaps you have a thing for profound metaphors, or richly furnished lists, or you share my fetish for killer verbs.
Perhaps no single sentence captures the essence of perfection, but a fragment of the archetypal form resides in all of them, too vast and multidimensional to fit all of it in a single line.
One may notice that the perfect sentence doesn't have an adverb, and then proclaims "don't use adverbs," which, to be fair, does nudge you in the direction of perfection.
Thinking of it as a single sentence might be misleading — perhaps it's better to imagine the archetypal form as a passage, perhaps a few paragraphs long: a stylistic summary.
This all feels nicely general. The rules all derive from a prototype that depends on personal, subjective taste. But once you figure out where lies the target you're taking aim at, you can objectively evaluate if rule makes the trajectory more accurate, or throws it off, or only sometimes hits the mark.
But what to do if old standbys like "avoid adjectives" or "don't be vague" fail you, because you've already internalized them, or simply care not where they lead? How do we make up our own arbitary rules to blindly adhere to?
What can you do if you know there's a gap between your efforts and your idol, but you just can't see what to do differently?
We can reverse engineer what led the common rules to exist in the first place. The value judgment is a component, but not the only one: also important is that they push you away from default phrases.
So, pay attention as you're writing to what phrasing you reach for first, and then as you're rereading, notice what rhetorical tropes keep cropping up. Then challenge yourself to avoid the old reliables.
(When I write, some of this comes instinctually. Half the time, after putting down an awfully normal-sounding sentence, I think to myself: how do I make this weird and unexpected? How do I twist the words around to surprise the reader? How do I increase entropy? And what is increased entropy but increased temperature — the key ingredient for annealing.)
Do you always find yourself stapling on a conjunction or relative clause? What happens if you're only allowed one connective per paragraph? Do you overload sentences with images and descriptors? What if you honed every sentence to each convey one image? You love em-dashes, don't you? (We all do). But how can you convey that same spontaneity and interconnection without it — can you wield exclamation marks or semicolons?
At a certain point, this becomes less of a quest for "rules" and more of a collection of writing exercises. But I always found it odd — artists do warm-up sketches, musicians practice scales and repetoires for hours, and yet there seems to be no culture of practice among writers. All our efforts go toward the final product all the time.
And from that angle, my advice might seem silly, even misguided. What good is a rewriting pass that butchers the text in service to rules you know aren't the end-all be all?
You can try annealing directly, unburdened by rules. Scan through the text, find the weakest, most ill-fitting lines, and then just mentally jiggle the syntax until you find a more tightly locked form. But at that point, this starts to sound like an ordinary process of rewriting. Dressed up in a cute metaphor, but relying on the same skill-bound judgment.
Could this reframing help? I couldn't tell you. Here's a confession: this essay isn't me describing a process I've attempted in any depth. It's a whole lot of me imagining someone trying to improve their process, and working backwards to think of what I'd tell them to do. I think following rules like those I listed at the start helped me, but perhaps my memory has been distorted by the years.
When I sit down and look at things I've published, so often I find there's little I want to change. I scan for weakest lines, but I find it all flows. I look for recurring tropes to avoid, but again, why butcher a well-woven text? I don't think this self-assuredness in one's writing quality extends to most of the people reading this (or if it does, chances are I'd read your stuff and privately think it shouldn't :p).
There's a phase transition, between a state of your writing being riddled with flaws such that you can't go a paragraph without finding something that needs improvement (and thus broad rules will flag those sentences with statistical regularity), to the state where further refinement must be subtle and hard to notice (you knock the atoms out of place but most just settle back into the same lattice). I think these practices could help the former toward the later, but is it much help for the latter? I don't know.
For years now, I've felt that I don't actual try when it comes to prose, not anymore. When writing Endless Stars, I broke my back refining every sentence. Every novel after that has been defined, at its start, by me lowering my standards, letting myself slouch, getting sloppy and casual with it. Of course, each time my standards inevitably crept ever higher. I think Hostile Takeover is superior in every area I struggled with in ES, save worldbuildling, and I think its prose, too, is competitive. But is it?
So of course I wonder, as anyone who's been working for years will, whether I've really gotten any better. My improvement was definite at the start; resting on my laurels has surely slowed it down, but even if I'm not giving it my all, I'm still getting better, right?
Reasoning monotonically, at a certain point, my casual efforts must eclipse what I once accomplished with strenuous effort. So I occasionally ask the very few people who've read my oldest works: is the new prose any better?
And so often, they can't answer. Apples and oranges. The styles are too different.
These days, I outline more stories than I write. And I find, even when I'm excited for the outline (cough, Aurora Moonrise), when I sit down and try to start writing, I get that famous blank page anxiety. I've written the better part of a million words, so how am I still struggling?
Even for a single writer, there is no definite article here, no the archetypal form. Different works have different archetypes. Even in a single work, finding a character's voice, figuring out how to differentiate dialogue and the menageries of perspective, each require a distinguishing archetype. I can map out a story from beginning to end, but mere plot and lore does not an archetype reveal.
There's a metaphor I love, and I never get tired of using it. All of my notes and planning, no matter how detailed, amounts to a sky darkling with clouds. A chill creeps up, but on its own, even cold below freezing isn't enough to turn water to ice.
Each snowflake is ice crystallized around a grain of dust; it can be the smallest detail, but it takes a detail.
There are other metaphors to use. Underbrush can pile up and dry out dangerously, but it takes a spark to start a forest fire. One friend is fond of describing writing as the process of tearing out a splinter lodged inside you. And so on.
The great flaw of perfection is that it is form without content. Every decision of a specific word over another closes off infinite possibilities, and somewhere in that infinity is something a little bit better. Through revision you wrestle the true, raw meaning you want to convey, drag it closer toward your chosen archetype. You can call prose the aftermath of the war between style and substance.
The best of us can hope for a strategic victory.