Introduction
We’ve been over this before: a scene is people doing stuff in a place with a twist. Last year’s article covered a simple tool for outlining, the scene formula. By determining where a plot point happens and who’s there then adding what they’re doing and why, you have created a well-formed scene: all the pieces are there.
In the skyscraper’s elevator shaft, Khalko climbs the pulley in pursuit of the masked figure, only to pummet on severed ropes.
Rather compressed, but all the pieces are there. So, is this a scene?
Intuitively not; barring dramatic, exception-proving circumstance, you won’t open a book and find a raw formula like this bracketed by fleurons. It’s missing something. This is a recipe, not a meal. So how do we get cooking?
First, you’ll find it instructive to take these formulae a few steps further. See Wheels Within Wheels for a fully-worked out demonstration in all its glorious weeds, but to recap: if a skilled writer tries to outline a fic, they won’t just rattle off formulae.
You could, but it won’t feel right or come naturally; instead of just slotting in people, stuff, and places as if playing mad libs, you’ll instead feel the impulse to start storytelling. You’ll add summaryish bits of connective tissue to explain how one scene sets up the next, or clarifying what specifically the characters are doing and how it furthers their goal.
If you’re truly experienced, you’ll already feel the siren song of details. When I first outlined Aurora Moonrise, by the time the second act’s outline was wrapping up, I didn’t just have a list of formulae with a few extra ornaments. No, characters and conflicts had become so clear to me that I needed to transcribe snatches of dialogue which evoked the point of the scene so much better than describing.
After all, why say “a perplexed comet taunts Aurora as they fight battle in its flooded crater”, when I can demonstrate the exact psychological faultlines Vela is going to probe while implicitly suggesting her confusion at what our half-comet protagonist even is? Worse, I have a keen ear for dialogue; if I come up with one punchy line, I’m already thinking of the antecedent lay-up or a quipy retort.
And I get carried away so easily. Aurora Moonrise was just the beginning. I keep writing these so-called “outlines”, and they only get longer; An Opaque Heart begins as clear summary, but by the epilogue, you’d be fooled into thinking you were reading chapters of an ordinary fic.
These results are clearly unwieldy; such lengthy passages might express the point better, but you can’t line up ten of them on a page in a nice bulleted list and see the flow of an arc in a single sweep of your eyes. Is it even an outline, if you’re writing several hundred words to describe what might end up as what, a few thousand?
If these indulgently long summaries don’t make better outlines, what good are they? Truth be told, I so often have no idea what to do with them. I keep producing them, because I’m good at it and people like reading them, so I find this a quick and rewarding way to get ideas out there.
I call them ‘semidrafts’. You might call them what they are — first drafts — but I dislike that, because my real first drafts look pretty much like my published drafts sans polish. Still, the conventional idea of a first draft is that you rewrite it to yield the second draft.
So obviously, that’s how you turn these overgrown outlines into something presentable: you rewrite them.
Maybe this comes easy to others — people so often talk about hating their first drafts, and needing to edit and edit to get something halfway decent. Bear with me, because this will start as a brag: I don’t have this issue. I’m pleased with my first drafts, and even with those off-the-cuff semidrafts — half of which literally began as discord posts that got far, far too long — I look at them and find them to be outright good writing.
But trust me, this is suffering from success: it means I can’t take the shortcut available to everyone else, guided by embarrassment and self-critique to revise again and again until the results meet my standards. (Maybe my standards are just low; my taste suggests this.)
Remember, the core principle of editing is first, do no harm. If I like my summaries, how do I know the expanded version is an improvement? I have this sense that “real” writing “ought” to be more detailed than a semidraft, but why? How and where must the expansion happen?
To making good something bad is (if not easy) a process with a clear sense of direction. Correct what’s wrong and add what’s obviously missing. But to make something good even better requires a keener discernment — and that’s why I think my “brag” demonstrates the true problem and my motivation here.
Solve this problem in the hardest case, and the rest comes easy.
The Five Components of Scene
Let’s dissect intuition until we can see every mechanism moving.
“Scene” and “summary” must lie along a spectrum, but a spectrum of what? Length, first of all: scenes are much longer. Length to read, yes, but also length to write. Scenes are harder, much harder, and this is why we care about the spectrum.
The thinnest summary is no fun at all, and the richest scenes become sublime. While I can produce detailed summaries nearly at will; the inspiration to write with true seriousness flickers like a dying flame. I weigh the cost/benefit, and a goal compels me: how can I write something scenic enough to pass muster for enough readers, yet summary enough I don’t need to wait in vain for the muse to strike?
So again, what is this a spectrum of? What is that extra length doing that lends it such an essential character? It’s not just one spectrum, really. One might call it a multidimensional space if the axes didn’t overlap and compound on each other. But I’m getting ahead of myself.
Detailing
First: scenes are more detailed. Years ago, in my first real essay, I discussed what this really means at some length. To obtain scene from summary, you simply “show, don’t tell.” Easy to say, but in practice, a deeply ensnaring topic to discuss — that single essay is over five thousand words and breaks the concept down into seven (7!) different ways being more detailed can improve writing.
Some of that bleeds into the things we’re about to cover in this essay, and I really don’t want to spent too long blathering about what “detailed” really means. You know it, I know it.
What I do what to comment on is how unhelpful this understanding of scene is.
A knight fights a bandit — let’s sceneify this. That means more detail, right? So let’s talk about their armor’s plates and patina, the positioning of their limbs, the expressions inflecting on their faces. Wait, no, let’s set the scene first. Is this happening in a fetid swamp? Let’s talk about what plants and animals decorate the background. Have we mentioned the dark and stormy weather?
I talk about sensory reification near-exclusively, but exposition is a whole other category of detail. Every piece of equipment on each fighter bears its own story. Can we recount what paths they’ve traveled, why they’re fighting, what they’re plotting?
But I’m preaching basics here; this is a well-known terrible way to write. It saps the drama, distracts the reader with filler and trivialities, it wastes time.
Still, skill issue, no? Sure, you need to add the right details in the right place, but everything in writing comes down to knowing how to do it right. Regardless of pitfalls, a scene must have more details than the summary, by definition. (No pun intended.)
Once you know what’s right and what’s wrong, you refine summary into scene by going through and swapping the vague bits with (good) detailed bits. Right?
I think this unhelpful. Sounds good, doesn’t work. In Outlines as Temporarily Embarrassed Drafts, I wrote:
The theory is that diligently following this rule, I could recursively flesh out a whole story. But that’s just theory; it’s never worked.
Moreover, what I find as I try to expand stuff like this is that this structure starts to really buckle with just one or two levels of recursion
When you think about it, this bulleted list just isn’t shaped like a story, and it becomes more and more noticeable as you inflate it with words.
How do you turn a stick figure into a painting? Do you render it pixel by pixel like you’re a wetware graphics card? No, you need a sense of form and anatomy. And when there’s anatomy problems, you just might need to do surgery.
Flowing
There’s something everyone talks about, but no one seems quite clear on how to do it. Your prose should flow. You don’t want it to be stilted and choppy; you want it to be tight and smooth.
To be fair, this is not a deep mystery: how prose flows is how it jumps from one idea to the next. Smooth flow means small gaps between ideas, tight flow means the ideas lead into each other.
Summarize, and you get the freedom to leap to conclusions without showing your work, to bring up whatever you need just out of expedience. When a writer summarizes, the reader lets them cook; they accept things with no more than a gesture, imagine a more realized whole between the lines.
Trying to close these gaps requires more detail (and thereby length). And not just in justifying connections step by step. What happens if it’s raining and you’re wearing shoes with worn down soles and the ground beneath you is ceramic tile? You slip. Conclusions depend on multiple premises, and sometimes these ideas aren’t related until that final stroke of logic. This is the fundamental problem of getting from A to B.
Smooth transitions require clever (sometimes obviously strained) segues, or plank after plank of pure bridging material.
But we skipped over something crucial here: what does it actually mean for two ideas to be “close” or “distant”? I won’t be rigorous; a sense of conceptual distance came free with your ability to understand language. But let’s at least coin some terms to talk about this.
Sentences are the unit of analysis here. Paragraph breaks allow for smoothly crossing larger gaps, and scene and chapter breaks even larger still, but we’re talking scenes here.
The gap between two sentences is the relational distance between them. What’s the relationship between them? For instance:
- If they share the same subjects, they’re related. The rain fell. It was cold. And it didn’t stop all day.
- A subject is also related to its components, and vice versa. The rain fell. The droplets splashed. The water hit me under the awning.
- A subject is related to its immediate (causal or logical) consequences. The rain fell. Puddles formed around the road. Streams ran out of overflowing puddles.
A subject is also related to things nearby in time or space (after the rain, a rainbow), and things that share properties with it (the rain descended and with it the excavation team), and so on. This all isn’t exhaustive (I’d argue there can even be a kind of stylistic relation formed between sentences that structally mirror each other and nothing else), but you get the idea I’m gesturing towards.
It’s tempting to dive further into the weeds — word order also matters (connections are tighter the physically closer they are, the end of a sentence binding tighter than the start), as does a kind of momentum (consistent relations bind tighter; it flows better to go consistently from left to right than to jump back and forth rhythmlessly), but I could be here all day. Let’s go broader.
Relational distance is fundamental to what scenes are — it’s people doing stuff in a specific place during a certain interval. Many writing manuals privilege that unity of time and space in their definition of scene, and it’s very important, but I hope you can see that it’s just a consequence of profluence as a guiding principle: telling a story by locking into the here and now is just a logical way to minimize relational distance, especially if you’re doing sensory reification.
Now, if profluence is so great, why don’t we flow smoothly from start to end? Is every scene break a dirty cheat or admission of defeat? Obviously not, an unending step-by-step elaboration would be an exhausting waste of time.
There’s an important relationship I saved until the end — abstraction, that telescopic shifting of a lens away from scenic detail to high summary. After all, everything connects with around six degrees of conceptual separations. It’s like pathfinding on wikipedia; the fastest way to go almost everywhere is to first go up.
And the easiest way to bridge two ideas is to gracefully zoom out. An old concept in the writing advice memeplex is the concept of “pyschic distance” (which might as well be its own axis of relational distance.)
Contrast “The temperature is -5C,” with “Shit! My snot’s freezin’ to an icicle right under my nose!” One of these is so much more tightly locked to a character’s perspective, and it’d be jarring to switch between them carelessly.
So much of writing comes down to rhythm. Just like the balancing of short sentences and long sentences, of description and dialogue, of intensity and breathing room, flow is its own game to master.
Short relational gaps are a good default, but a sudden jolt can inject so much life. A choppy passage can express what meticulous care cannot. And then there’s the modulation of abstraction itself, for context, for negative space — in short, every scene needs moments of summary.
But the danger of overemphasizing flow isn’t tedium alone — because flow isn’t simply stringing sentences together. Proper profluence flows toward something, lest you court unstructured rambling.
Pointing
How do you fix a rambling run-on sentence? You split it up, add a full stop in the middle. You might think the metaphorical full stop of a scene is the scene break, and maybe it is, but that then raises the question: what is the comma?
In practice, scenes are a few thousand words long — which is ample time to trail off and lose direction. So, what keeps you on track?
The two tools I’ve picked for analyzing this are “punchlines” and “icons”.
A punchline isn’t necessarily funny, it’s a line that hits you. Makes you go “ooh.” Importantly, punchlines aren’t freestanding, they need setup to land, so the proceeding passage is structured to prime the reader for the payoff. Building to the mini-conclusion, the reveal, the point.
Elsewhere, I’ve defined storytelling telling as the juxtaposition of dramatic and sensory detail — and punchlines are all drama.
An icon, then, is the complementary notion. It’s sensory, often visual — they are cool details, specific images that excite and evoke. But I can imagine icons that set up, icons that are intermediate of the true punchline.
The purpose of an icon is to inspire a sense of cinema. An icon is a particular shot with great composition. It’s the sun rising behind a charging soldier, it’s the ironic remark right as the sirens go off, it’s groping around in the dark mud for dropped key and finding a snake first.
Icons serve a visual outline (imagine muting a movie and trying to follow along; you’ll struggle, but how much?), and require a specific set up (just as a great shot needs the right lighting, the right angle), or otherwise must yield immediate consequences and reaction. It’s good form to give the reader something to imagine, even in the midst of dialogue or introspection — even if it’s just metaphor.
When writing semidrafts, dialogue comes to me faster than action. (I think in terms of perspective and motivation, and dialogue is distilled character.) Even when outlining more serious fics, sometimes a great exchange comes to me, without no stage directions or set dressing. (So many big conversations in later chapters of Hostile Takeover began life this way.)
Icons are a practical answer to this particular problem of mine, of needing to stitch in description in a separate pass from dialogue.
But we’re dissecting every organ here, no kidney stones unturned. So why do I do this? Why bother weaving in description?
Layering
Scenes require a sense of place — but why not lay out the set in one big descriptive “SCENE:” block at the beginning? It’s good enough for scripts, why not for us?
Even a tightly flowing, step-by-step account of pure conversation will include a little body language, emphasizing and reacting to the utterances. But how much of this do you need? If you just want to flow, statement-response chains can have some of the tightest flow out there.
Maybe that notion of variety is enough to save us — you need just enough description to break up the monotony of the dialogue. But even that is its own kind of monotony; adding any text purely out of formal consideration always invites filler.
So often, your first impulse in detailing a dialogue is to harmonize it. Character says something angry, so they do it along with angry body language and angry tone. “No,” she says, shaking her head with a frown.
It strikes me as dreadfully boring, and growing wary of this, I now think instead of my additions as being an exercise in counterpoint.
Musical counterpoint is a rich tradition, but its key aim is to produce melodic lines that flow independently, maintaining distinct identities without overmuch dissonance. I never studied the art deeply, but basic tricks are easy to pick up, such as contrary motion — one line goes up, the other down — or staggering notes so that each line draws the listerner’s attention on different beats, rather than being locked in concert.
Counterpoint doesn’t mean ‘do the opposite’, it means no voice is subordinate to another. Each is a compelling melody, but in composition they interleave into something altogether more entrancing.
It translates well to literature; writing gurus before me have talked about plate-spinning. What is the literary equivalent of a voice line, though? What are the registers of prose — is it only description and dialogue?
Call a throughline any part of a scene that operates by its own consistent logic, bearing a characteristic voice, in contrast to other throughlines. A monologue might have only a single throughline (though there are of course nuances). A dialogue has two interfering throughlines, assuming there’s any conflict or drama to the discussion.
Geometrically, two lines define a plane, so let’s identify each pair of throughlines with a layer. In a simple dialogue, a character arguing with another forms a single layer.
The archetypal scene, through this lens, operates on at least three layers. The three most common throughlines are intent, action, and obstacle.
- Intent–Action: A character comes into the scene wanting something, but what can they do to get it? What actions are possible, which one is best?
- Action–Obstacle: Whatever actions they select meet with resistance. So what complications result? How do the characters evolve their approaches to overcome each other?
- Obstacle–Intent: Finally, seeing the opposition they’re up against, a character may reconsider what they even want, through mind change or mission creep.
You can see how friction on all three layers enriches. White room dialogue has only a single layer; add in some description and you have one more, and if the narrator is introspecting and plotting their moves throughout it all? Now we’re cooking something that feels like a proper scene.
The interactions of thinking versus doing versus saying is just one possibility. If there’s no dialogue, if it’s a fight scene, then a mere blow by blow occupies one layer. Focus on the tactics and strategy, and you win another layer — but what about the environment?
Place is a key part of every scene, and it should be character in its own right. How does it present friction and opportunity for the protagonist and antagonist? How does it shift around them? I claimed each throughline has a “voice”, but the voice of a setting is better understood as its atmosphere.
Several characters might each have their own throughline, or more often, they align into factions. But I can’t complete this taxonomy; the specific anatomy is distinct with each scene.
A general rule of thumb seems poignant, though: three strikes me as the magic number. Less than that, it’s sadly flat. Too much more than that, and the number of interactions grows quadratically out of hand.
To render a good scene, aim for a layered trilogue.
Back when I first contemplated writing this essay, this exercise in literary counterpoint was initially its only topic. “Putting Flesh to Bone” would be my guide for bringing the skeletons of scenes to life — I just had to figure out if my experimental results were alive.
That essay, as first conceived, didn’t really come together; I hadn’t yet shape-rotated the topic to my satisfaction.
But, after three thousand words, we’ve just about arrived at this original starting point. So, how do you expand a scene skeleton? If the skeleton is your only guide, if you stretch gaunt skin over the frame, the result will be just as dead.
No, try putting flesh on the wrong bone.
Riddling
As this section draws to a close, it’s fitting to round off our model with a discussion of how to end scenes. Punchlines are the commas, so what could the full-stop be but the final climatic punchline of them all?
An easy misstep is to write scenes that peak early then peter out chasing their flow and resolution. Sometimes there’s no proper final punch at all, simply ending. There’s a craft to writing cadences that resound — but all that’s outside the scope of this essay.
No, the last piece to the puzzle that is scenecraft isn’t finishing them, but finishing them. Not the first draft, but the final draft. Everything we’ve covered so far lays it all out, but the last piece is what ties it all together.
Scenes begin with a single line, a nice opening hook. From this detail spring other sentences, a relational flow. But this flow needs to go somewhere, drive home a point. And that route forms the first throughline, but it isn’t the only one; layers of complementary lines are drawn out from the proceding details to play off the first. And then…
What’s the point of all this?
Here’s a question: I am big and hot — what am I?
Did you answer right? Did it feel like an answer, or a guess? Does this count as a riddle? It parses, sure; it’s a valid answer to the question. But so is “a huge campire”, or “plus size lingerie model”, or “viral yet controversial tweet.”
Consider instead:
With pointed fangs and in plain sight, my bloodless victims are bound by my bite — what am I?
The answer?
And can’t you just feel that specificity and contradiction constraining this to have only one answer that isn’t profoundly tortured? How hard would it be to remove or change something without breaking this delicate logic? It’s precision without redundancy.
You see hints of this in all the other principles — if your details matter, if your flow is seamless, if your punchlines have setup, if your throughlines act in true counterpoint, then you’re already grasping at the final principle.
But this, I contend, is what it truly means for a scene to be fully rendered, to have high fidelity, to be finished. Not just long, but each piece arranged to advance with all the inevitability of a riddle.