The question got dropped on my doorstep, finally.

I can’t ignore it anymore. It’s everywhere. Let’s talk.

Let me begin by demonstrating where some well-meaning folk are in denial and where I am not.

Yes, a model can produce a beautiful sentence. Yes, output-for-output, generated and authored prose can be indistinguishable on the page. I grant this fully so that we’re sitting in a realm of reality when I tell you why this is okay and why your work is not destined to irrelevance in the face of some super-machine author/illustrator/artist meta.

So, if the page can’t always tell you, then where is the difference? Between you and generated prose/visual?

The Relationship with the Work

Not in the “this means something to me” sense but in the sense that you, present in every step of its probably rather lengthy creation, are fluent in it in the way no prompter could possibly be.

Your authorship? That exists in a rejected alternative.

In case you weren’t aware of how this works, AI isn’t true intelligence. It’s token-based on probability. This is why generated writing tends to have this recognisable cadence you’ve all come to know and loathe.

It’s become the green carnation of the writing world. Em (—) dashes. Not X, but Y.

Because, when these models were authored—ironic and we love this—the work that was used to inform its capacity for competent, advanced prose was the work of competent, grammatically correct and sometimes un-adventurous writing that has been condensed now into a pattern-recognition nightmare. One that informs any machine how to generate, edit and translate text.

So, now you see the context for the reasonable argument that sometimes competent writers look like AI-prompters. Even cadence, punchy declaratives. Sometimes an instinct to over-explain for a demographic. Things you may have learned or been taught to do that now get or have gotten you scrutiny.

A model optimises for that center and you, artist, are defined by where you leave the center.

I’m not here to console you—after all, I’ve just admitted AI can craft a lovely sentence. But what I’m going to explain will, hopefully, console you anyway.

I’m a transmedia creator. Which means that I work in more than one sphere of media. I’m not just a writer, I’m an illustrator and 3D artist, full pipeline. I’m also multi-lingual and that gives my english prose, I think, a foreign strangeness I should truly lean into more, not less.

What this experience has given me is a lens into how the minds of the many people I’ve worked with think about their work. And how they talk about it.

This is where I was able to realise something. That when you make it, you thought about that decision. When you prompt it, you’re handing authorship over.

And therefore, in making it and swerving from probability-based decision making—if you have any form of inner monologue with yourself—you have just nurtured your ability to talk about your work in ways that someone simply trusting an output just cannot do.

Your fluency in your own work under questioning is where true-authorship lives.

This is also why many argue—and I would strongly agree—that generated work lacks “soul.” Because soul is strangeness. It’s chaos. It’s emotional and unpredictable decision-making that a number-as-probability generator simply cannot do because it is constantly optimising for the “center”. The most likely. The most “what would very skilled and successful author x, y and z agree is the most appropriate and correct word to use here?”

And people want soul. This is art. Experiencing the world through someone else’s eyes is exclusively possible through the consumption of it.

Some media, I would also argue, may not be “AI-generated” but is crafted so precisely as to offend no-one, appeal to the most people possible, be the most comprehensible to the most people and be as un-challenging as possible that it is doing the same probability-based generation that a model does. Only it’s a person optimising for eyeballs (see: money.)

I’m not making an argument on what’s worse. Not at all. I’m attempting a dissemination of how to define authored-versus-generated writing so genuinely creative people can stop feeling worthless while we’re all trying to figure out what our relationship with this is going to be in twenty years.

So. The author can reconstruct the why, defend it, have it argued with, because the choice was real and made against options. And the generator cannot go past the surface, because there was no why; there was only probability.

Great, but for some of us, that wasn’t enough to feel better.

Your Body of Work

Fine, you may say. But someone could fake the whole apparatus!

I find that to fake sustained authorship—that is, holding a voice across a body of work, a world built below the waterline and a defensible, demonstrated why across thousands of choices—costs more than authorship.

The sustained forgery theory requires an accused artist to be more impressive than the person they’re accused of not being. Which is why I posit one looks at the body, not the sentence.

One artifact can be doubted; a corpus held together—authored—by one consistent intelligence cannot be faked into existence cheaply.

That’s you. Hopefully.

You can’t fake your own repeated aesthetic obsession. Not cheaply. And your ability to intelligently talk about these engines that drive you every decision, often down to word-level, is what makes your work authored. Valuable. Yours.

Let me try to demonstrate what that could look like to you.

I received critique on possibly having overwritten some prose with “odd” language choices that was argued to be incorrect or over-the-top literary varnish on a mundane detail that didn’t need it. “Showing-off” at worst.

An example; “Why did you choose the word ‘beset’ here?” (Context: a man’s face as “beset by deepening lines.”)

My pushback, close to verbatim, was in saying that the mood of a world lives inside how small details are described. That "beset" is predatory—it carries yielding, or fighting (and possibly flagging behind.) I'd called it "elevated," but isn't that what literary writing is? Taking what's "just" anything and transfiguring it into something aesthetic?

I wasn’t defending one pretty word. I was defending a principle: that the register has to go all the way down. That the same sensibility describing the world’s main menace and antagonist has to inflect a man's face, or the world doesn't have texture.

I don’t believe any model in the world could curate its own aesthetic. That’s not what models are built for.

That’s what you’re built for.

You’re You and You Use Tools

What a machine can’t take isn’t speed or polish. The line here, that I’m drawing, is not ‘did a tool touch this’—tools touch everything. We research with them, we draft bullet-point notes and outlines with them. The Authors Guild itself carves out brainstorming, structuring, grammar, translation and only draws the line at generating the text.

The question here is not what assisted you, it’s accountability. An author/artist who can defend every sentence wrote it, whatever they used to get unstuck.

An ‘author’ who can’t, didn’t write it, no matter what produced it.

The more generated text there is out there, the more an authored swerve stands out. Volume of average makes the strange (we want strange) more visible, not less.

Generating is cheap. Prompting is cheap.

Sustaining your authored vision across your body of work is expensive. You know every tributary, though. You can defend those waters because for every most-likely-probability turn you could have taken and didn’t, you built on your body of work with a chaotic, passionate, human vision where no model could fake experience. That’s where art lives. Where its soul lives.

Let that silhouette your work’s unique spirit. Its edges and form. AI sands off every corner and makes everything safe. Competent. Structurally perfect and entirely boring.

Kind of like Hollywood.

Interesting.

Fair Winds,
D.S.

There’s a feeling you know.

You’re working on the subplot. It’s doing Important Things, you tell yourself. Delivering information, positioning characters, enabling the scene you actually want to write. You’ve justified its existence. You’ve revised it three times. And still, every time you sit down to work on it, your hands slow. The prose comes out wooden. You find yourself checking email, refilling your coffee, doing anything except pushing through the next paragraph.

That would be me. I spent two weeks in this exact state. A subplot in the manuscript was doing three jobs at once: revealing a secondary character’s divided loyalties, giving my protagonist critical intelligence before a major scene, and providing him resources to participate meaningfully in what came next. On paper, it was essential. Every thread it touched depended on it.

I kept rewriting it. I was changing who was involved, adjusting the pacing. Adding justification, cutting justification. Moving it earlier, then later, then back. Nothing helped and it was driving me bloody mad because if I could just get this dialed in I could finally continue. The subplot sat in my manuscript. Something about it utterly lifeless.

The Diagnostic

It was desperation to move it along that led me to first seriously consider cutting it. What would I lose? Truly. Turns out, actually, not a whole lot of anything. The subplot was contrivance, not character.

Contrivance serves plot mechanics. It exists because you need X to happen before Y can happen. The character does something because you, the author, require them to do it. The sequence of events is logical. It might even be terribly clever. But it doesn’t emerge from who these people are or what they actually want but rather it emerges from your outline.

Character-driven plot is different. Things happen because of who people are, what they want, and how they’d realistically pursue it. The sequence of events might be messier. It might not hit your structural beats as cleanly. But it breathes. Readers can feel the difference even if they can’t name it.

The test isn’t “is this subplot necessary?” The test is: does this feel like something these people would do, or something I’m making them do?

My subplot utterly failed that test. The mechanics were sound. The character motivations were thin. I was pushing pieces around a board instead of following people through their lives.

So I cut it.

Didn’t even trim it. Just deleted the scenes, removed the thread from surrounding chapters, accepted that three jobs now had no home.

And here’s the thing about cutting something that isn’t working: when it’s the right call, you feel it immediately. Oh, that relief I felt. The story started moving again. The energy that had been trapped in that dead subplot flooded back into the manuscript.

What Emerged

The scene I needed to write was simple: my captain summoning his second lieutenant to brief him before a critical social engagement. The failed subplot had been loading this moment with external mechanics, logistics, information drops, resource transfers, loyalty tests. All the weight was in what got exchanged.

When I stripped that away, the weight shifted to who was in the room.

The lieutenant who enters isn’t delivering plot information. He’s a man whose uniform is always immaculate. He’s not vain, per se, but a lifetime of being watched and measured and found wanting by standards most men would never comprehend makes him cripplingly self-aware. He wears his perfection like armour.

The captain isn’t receiving a briefing. He’s bracing for judgment. His lieutenant is clever enough, observant enough. If anyone on this ship could see the cracks, it’s him.

The blade doesn’t come.

Instead, the lieutenant pivots. Becomes genuinely useful. Offers expertise freely, without positioning for advantage, because someone finally stopped treating him as a threat. And the captain feels something loosen in his chest—not trust, not quite, but the specific relief of a man who had braced for a blow that didn’t land.

Then the First Lieutenant arrives. The old one. The loyal one. And he reads the situation instantly:

“You’re taking Gore.” It didn’t sound like a question. Just the flat recognition of a man watching his captain choose a weapon he couldn’t provide.

The scene ends with two weights in the captain’s chest instead of one. He made the right choice. He knows that. It doesn’t feel like the right choice. It feels like a small betrayal.

None of this was in my outline. None of it could have emerged from the contrived subplot I’d been protecting. It happened because I stopped asking “what does the plot need?” and started asking “what would these people actually do in this room?”

The information still got delivered. The relationships still advanced. But now they advanced through character truth instead of mechanical necessity.

The Principle

Kill-your-darlings advice assumes you’re cutting something beloved. Something precious you’ve grown too attached to see clearly. That’s real. This happens.

But this is different. When cutting feels like relief, you weren’t killing a darling but were removing an obstruction you’d mistaken for load-bearing structure.

Not everything that feels essential is essential. Sometimes what feels essential is just needlessly, stupidly complicated. You’ve invested so much work justifying its existence that you’ve convinced yourself of its necessity.

Or perhaps you wasted so many hours writing it you hated the idea of tossing the work.

The story knows when you’re forcing it. That wooden feeling, that resistance, that sense of a Sisyphean uphill every time you think you’ve solved it—these aren’t signs you need to work harder. Something in the machinery is binding.

When you’re constantly justifying why something has to happen instead of simply watching it happen ask whether it’s earning its place through character truth or through mechanical necessity.

The former will carry weight. The latter will make you check your email twenty-five times an hour…

Something that helped my manuscript work flow again after weeks of stagnation, delivered with brevity. Hope this gives you permission to just cut.

Subscribe now

Fair winds,
—D. S. Black

Function vs. Haunting

There’s a distinction I use when building characters that separates the ones who function from the ones who haunt.

What they want is the surface. The conscious goal. The thing they’d tell you if you asked. Promotion. Survival. Revenge. Love. It’s legible, articulable, and usually drives the plot.

What they’re looking for is beneath. The need they can’t name—often invisible even to themselves. It’s not what they’re chasing. It’s what would still be missing if they caught it.

The protagonist of my novel The Reply wants to survive. Wants to maintain command of his ship. Wants recognition from an Admiralty that despises his peculiar gifts. These are his goals. They drive his actions. A lesser version of the character could run on these wants alone and be functional—he’d have clear motivation, generate conflict, pursue objectives.

But Somerset is looking for something else. Something he’d never say aloud because he doesn’t have language for it.

He’s looking to be claimed.

Claiming isn’t like being needed or merely used and employed. Claimed—by something vast enough to see him fully and want him anyway. The sea that hunts him. The officer who mirrors him. The divine attention that might destroy him but would at least know him first.

That’s why the plot can resolve and the character can’t. You can give Somerset everything he wants—command, recognition, survival—and he’d still be looking. The want is achievable. What he’s looking for is a hole in the shape of God.

This distinction is the difference between characters you remember a week after finishing the book and characters who take up permanent residence in your mind. Function versus haunt.

Most contemporary fiction has forgotten the difference.

Characters want things. Clear things. The plot provides obstacles. The climax resolves the wanting. Everyone goes home. The problem is that these characters only exist on the surface—because their creators do too.

You can’t write the looking-for if you’ve never asked yourself what you’re looking for. And that question requires a kind of interiority that’s becoming increasingly rare.

The Flattening

You can’t write what you can’t access in yourself.

This isn’t mysticism. The looking-for—that unnameable need beneath the conscious want—has to come from somewhere. You can’t invent it from nothing. You recognize it. You find it in yourself first, then give it to the character.

Which means the craft failure has a source: writers who’ve never asked themselves the question.

Not out of refusal. It’s not a matter of will not, but cannot.

There’s a term from psychology that I very much like: interoception. The awareness of internal states. Hunger, heartbeat, the texture of your own unease. The capacity to notice what’s happening inside you before you name it, before you explain it, before you translate it into language someone else can understand.

This capacity can be developed. It can also atrophy.

A culture that can't sit still, can't be alone, can't tolerate ten minutes without stimulus, produces people with diminished access to their own interiours. If you've never been quiet enough to notice the difference between what you want and what you're looking for, you can't write characters who carry that distinction. You'll write the surface. Legible wants. Achievable goals. Flat.

I grew up with European parents in America. Spanish, German and Polish. High-context communication, where what matters lives in subtext, in implication, in what remains unsaid. I learned early what it costs when the culture around you can't hear le non-dit. Everything must be stated. Subtext is "unclear." Implication is "poor communication." You're forced to translate yourself into explicit language—and something dies in the translation.

American communication has become pathologically low-context. This isn’t an accent or a dialect but a flattening of the entire register in which complex interiority can be expressed. Characters in American fiction explain their feelings. They announce their motivations. They narrate their growth. They do this because their writers do this—because the culture has forgotten that anything can be communicated without being said aloud.

The result is fiction that functions like a workplace email. Everything important is stated. Nothing is left for the reader to feel into. The text doesn’t trust you, because the writer has forgotten that trust is possible.

This is the disease. The craft failure is a symptom.

Writers who’ve lost access to their own depths produce characters who don’t have depths to access. The looking-for requires interiority. Interiority requires silence. And silence has become so intolerable.

The Marvel Problem

Let me be specific about what flat characterisation looks like at scale.

Marvel villains want things. Clear things. Legible things. Thanos wants to erase half the universe. Killmonger wants to arm oppressed people worldwide. Hela wants to rule Asgard. The goals are stated explicitly, often in monologue. The heroes oppose them. The conflict resolves through combat. Everyone goes home.

This isn’t bad storytelling. It’s functional storytelling. It generates conflict, sustains a plot, delivers spectacle. The machine works.

But ask a different question: what are these villains looking for?

Not what they want. What need would still be unmet if they got everything they’re chasing?

The answer, in most cases, is that the question doesn’t apply. There’s no beneath. Thanos wants the snap. That’s it. He’s not looking for anything underneath the goal—no unnamed wound, no inarticulable absence, no hole shaped like answer. He’s a function dressed as a character. A plot obstacle with aesthetic flair.

The most Marvel can manage is making villains sympathetic. Killmonger has a sad backstory. Thanos believes he’s righteous. The films work hard to make you understand why they want what they want. This is mistaken for depth.

It isn’t. Understanding someone’s motivation isn’t the same as complexity. A character with a legible backstory explaining a legible goal is still flat—just flat with context. Sympathy is not interiority. Explanation is not the looking-for.

Compare Hannibal Lecter.

Hannibal wants things—escape, fine dining, freedom from tedious people. But he’s looking for something else entirely: a mind capable of meeting his. Clarice doesn’t just oppose him or help him. She sees him. That’s why he’s obsessed with her. That’s why the relationship is the engine of everything. You can’t resolve that by catching him. You can’t defeat recognition.

Or consider Daniel Plainview in There Will Be Blood. He wants oil, money, victory over his competitors. He gets all of it. The film ends with him alone in a mansion, having achieved everything he ever chased, and he’s more hollow than when he started. Because what he was looking for—connection he couldn’t admit he needed, a son who’d see him as human, some evidence that his existence mattered beyond accumulation—was never available through the goals he pursued. The want and the looking-for were pointing in opposite directions. That’s why the film is a tragedy and not a success story.

Marvel doesn’t make tragedies. It makes conflict-resolution machines. Efficient, satisfying, forgettable.

The audience gets what it’s trained to expect: problems with solutions. Wants that can be thwarted. Villains who function as obstacles and then stop functioning when the obstacle is removed.

This is what flat characterisation looks like when it has a billion-dollar budget. The spectacle distracts from the absence. But the absence is still there—that hollow space where the looking-for should be. You feel it in how quickly the films evaporate from memory. You saw it, you enjoyed it, you couldn’t tell me what Malekith wanted if your life depended on it.

Characters built only from wants are disposable. The looking-for is what makes them permanent.

You Write What You Can Embody

Here’s the uncomfortable truth: you can only write what you can access in yourself.

Not what you’ve done—what you can feel the shape of. What you can find a path toward, even if you’ve never walked it. The interiority has to exist in you before you can loan it to a character. You can’t fake depth. You can only recognise it.

I write men of violence because I've held violence in my hands. Not theoretically. Not from research. I've made choices in rooms where the wrong word meant consequences I'd have to live inside forever. I've been the calm one when calm was the only thing between a friend and something I can't name here.

My characters are contained because I am contained—and containment is not absence. The people who’ve called me cold, robotic, “Spock” (I’ll take this one as a compliment) , have never seen what I’m holding. They see the stillness and assume the stillness is all there is. They mistake the lid for an empty vessel.

Meanwhile, inside: a furnace. Spiraling. Emotions so strong they’d be illegible if I let them out unfiltered. So I don’t. I learned early that the world isn’t equipped to receive what I actually am. You adapt or you break. I adapted.

My characters know this. Somerset performs control while drowning. Origen processes trauma through millennia of pattern recognition because feeling it directly would annihilate him. Fressange aestheticises war because beauty is the only container that can hold what he's seen. They're not me. But I didn't invent their psychologies. I recognised them. They were already in me, waiting for names.

This is what I mean by access. Not autobiography. Resonance. The ability to feel the shape of an experience from the inside, even if the details differ.

I can write a man who loves his ship like a body because I know what it is to love something that can’t love you back. I can write cosmic horror because the numinous invades my quietest moments uninvited—the vertigo of deep time, the terror of a universe that owes me nothing and will continue without me. I can write the ache of men built for wars that never came because I know what it is to carry capacity that has no outlet. To be made for demands that never arrive.

If you’ve never been quiet enough to hear what you’re actually looking for—beneath the goals, beneath the plans, beneath the story you tell yourself about your own wanting—you can’t write characters who carry that weight. You’ll write wants. Legible, achievable, flat.

The question isn’t whether you’ve suffered enough to write deep characters. Suffering doesn’t automatically produce interiority. Plenty of people suffer and learn nothing about themselves.

The question is whether you’ve sat with yourself. Whether you’ve tolerated the silence long enough to notice the difference between what you say you want and what you’re actually looking for. Whether you’ve felt the shape of your own unnamed needs without rushing to name them, fix them, medicate them, scroll them into oblivion.

Most people would rather do anything than sit in that room.

And so they write characters who’ve never been in that room either. Flat people producing flat people, all the way down.

The Practice

This isn’t mysticism. It’s craft prerequisite. And like any craft prerequisite, it can be practiced.

The exercise is simple. The execution is not.

Sit with a character. Not their plot function. Not their role in the story. Them. Ask what they want. Write it down. Be specific—not “happiness” but the actual thing they’d reach for. Promotion. Revenge. The woman in the blue dress. The ship with their name on the commission.

Then ask: what would still be missing if they got it?

That’s the looking-for. The thing they can’t name. The ache that won’t resolve even if every conscious goal is achieved.

Somerset gets command. Gets recognition. Gets everything he says he wants. And he'd still be looking. Because what he's looking for is being claimed by something that sees him—and institutional success can't provide that. Only the sea can. Only Daud can. The want is achievable. The looking-for requires something that can't be pursued, only encountered.

If you do this exercise and come up empty—if the character only has wants, no looking-for—you’ve diagnosed the problem. The character is flat. Not because you made a craft error, but because you reached into yourself for the deeper layer and found nothing to draw from.

Which means the practice isn’t really about characters. It’s about you.

When did you last sit in silence long enough to notice what you’re looking for? Not what you want—what you’re looking for. The need beneath the goal. The ache that wouldn’t resolve even if you got everything you’re chasing.

If you can’t answer, your characters can’t either.

The practice is simple: stop. Be quiet. Be alone. Notice what arises when there’s nothing to react to, nothing to consume, nothing to distract. The discomfort that emerges isn’t the enemy. It’s the material.

Most writers would rather read another craft book. Watch another video essay. Collect another technique. Anything but sit in the room with themselves and notice what’s actually there.

I’m not telling you to suffer. I’m not telling you to excavate trauma. I’m telling you to pay attention. To develop the capacity to feel the texture of your own wanting without immediately naming it, fixing it, optimising it into a goal.

The looking-for lives in the space before language. You have to be willing to stay there long enough to feel its shape.

That’s the practice. There’s no shortcut.

The Death of Nuance Is a Choice

So. Flat characters come from flat people.

Not stupid or untalented people, mind. People who’ve lost access to their own depths—or never developed it—because the culture they swim in doesn’t require it and at worst actively discourages it.

You can get by without interiority. You can publish, produce, profit. The market doesn’t demand complexity. It barely tolerates it! Audiences trained on conflict-resolution machines will accept conflict-resolution machines. The feedback loop closes. Everyone gets what they expect. Nothing haunts anyone.

But the work that lasts—the characters that take up permanent residence in the mind—comes from writers who’ve done the harder thing. Who’ve sat in silence. Who’ve asked themselves what they’re looking for and stayed with the discomfort of not knowing.

I write men who were made for worlds that demanded everything because I understand the particular grief of being made for demands that never come. The soul built for storm, landlocked. The capacity for valor with no war to spend it on.The modern world didn't eliminate the capacity. It eliminated the demand. And capacity without demand becomes a kind of rot.

I languish in a demi-solde of modernity—half-pay, half-life, waiting for orders that won’t arrive from institutions that no longer remember what they’re for. I suspect I'm not the only one. The reenactors know. The wargamers know. Anyone who's ever felt overbuilt for the life they're living knows.

That’s not a complaint (although permit me some.) It’s material. For art. Because aren’t we creatives just the best at turning suffering into beauty?

The ache of wanting to be tested and never being tested. The grief of carrying capacity that rusts from disuse. The looking-for that can’t be satisfied by comfort, safety, the padded corners of a world designed to demand nothing of anyone.

Most writers have never examined this in themselves because it’s not comfortable. It doesn’t fit the therapeutic model where all desires are processed toward resolution. Some desires don’t resolve. Some needs can’t be met by the world as it is. Sitting with that—without numbing it, naming it into submission, or scrolling it into background noise—is the work.

If you’ve never felt the shape of that, you can’t write characters who carry it. You’ll write people who want things. Achievable things. Legible things. Things that can be obtained and then the story ends.

You won’t write the ache.

The best fiction isn’t written by people who’ve suffered most. It’s written by people who’ve stayed in the room with whatever they carry. Who’ve refused the easy exit. Who’ve let the silence get loud enough to hear what’s underneath.

The death of nuance isn’t inevitable. It’s a choice—made daily, by writers who won’t sit still, by audiences who won’t tolerate ambiguity, by a culture that’s forgotten that some things can only be communicated in silence.

You can choose differently.

But you have to be willing to stay in the room.

I write because there’s nowhere else for what I am to go. The capacity built for storms, spent on sentences. The valor that would have been spent on battlefields, transmuted into characters who get to live in worlds that still demand everything.

It’s not enough. It’s never enough.

But it’s the only legitimate outlet I’ve found.

For more craft essays and psychological character design navel-gazing, subscribe for posts every Tuesday.

Subscribe now

Fair winds,
D. S. B.

Everyone thinks they’re empathetic.

Ask a writer if they understand people different from themselves and they’ll say yes. Of course. That’s the job. They’ll tell you they care about perspectives outside their own, that they believe in nuance, that they reject simple binaries of good and evil.

They’re usually lying. Not deliberately—they believe it. But performed empathy can be a shield from truly understanding the mechanisms behind rendering complex psychology on the page.

Here’s what I mean: claiming you understand people who think differently costs nothing. It’s an identity badge, a way to signal sophistication without doing the actual work. You can call yourself radically empathetic while never once inhabiting a worldview that genuinely threatens your own.

The tell is always in the writing.

If your antagonist exists only to be wrong, that is, to be defeated and confirm the reader’s existing moral universe, you haven’t written a character. You’ve written a scarecrow stuffed with everything you despise, propped up so your protagonist can knock it down.

Scarecrows don’t reveal anything about the villain. They reveal the author. They say: I don’t understand people who disagree with me. I’ve never tried. I don’t intend to start.

The Method Acting Frame

Stanislavski’s “magic if” is usually taught to actors. It belongs to writers, too.

I discuss the extensive overlap in another post:

The technique is simple: you don’t observe the character from outside. You don’t describe what they do and assign reasons for it. You ask yourself, instead, If I were this person, with this history, in this situation, what would I do?

Not what would a villain do. What would I do, if I had lived their life.

This is the difference between watching a character and inhabiting one. Most writers watch. They describe behavior, assign motivations, move figures through plot like chess pieces. The character does cruel things because the story needs cruelty. The character wants power because wanting power is what antagonists do. It’s all mechanical. You can see the author’s hand on every lever.

Method writing requires you to disappear into the logic. To find the internal coherence that makes choices feel inevitable from inside the skull. Not justified. Not excused. Inevitable. The character couldn’t have done otherwise, because this is who they are, and you know that because you’ve been them.

This is uncomfortable. It means genuinely understanding why someone would do things you find repugnant. You have to find the version of yourself that could make that choice. The version that exists under different pressures, different wounds, different circumstances.

This is generally understood but recently conflated with “if you write this you must be this” illiteracy. But that did come from somewhere and it’s fascinating to see that actors, enacting the same technique, are understood to be playing a role, where writers are—increasingly—not.

When I write institutional antagonists, I can’t make them stupid. I can’t make them cartoonishly corrupt. I have to ask: why would I stay loyal to a system I knew was broken? And the answer is always human. Because I built my identity inside it. Because leaving would mean admitting my life was wasted. Because the structure gives me purpose and status I couldn’t find elsewhere. Because I’m afraid of who I am without it.

That’s not villainy. That would only be me albiet under different pressures.

Most writing stays shallow because this work is genuinely hard. It requires psychological risk from the author. You have to touch the parts of yourself that could become the thing you fear. And most people would rather write scarecrows than look that closely.

The Psychology Gap

Most writers don’t have a functional model of why people do things.

They work from types. Surface behavior. Tropes inherited from other fiction. Their villains are cruel because villains are cruel. Their heroes are brave because heroes are brave. The psychology goes exactly one layer deep, which is to say it doesn’t go anywhere at all.

I came to writing through intelligence analysis. Specifically, the part of the job that requires you to model how people think, what they want, and what they’ll do next. You learn fast that humans don’t operate on logic. They operate on attachment, insecuriy and wounds they don’t examine. You learn that the difference between instrumental aggression and hostile aggression changes everything about how someone behaves. You learn that shame drives more destruction than guilt ever could, because guilt says I did something bad and shame says I am bad. Guilt can be repaired. Shame has to be defended.

This isn’t all just academia. When you understand attachment theory, you understand why your character clings to someone who hurts them. When you understand defense mechanisms, you can write denial that feels lived-in rather than convenient for the plot. When you understand narcissistic wounding, you can write a villain whose cruelty makes sense. Not excusable, never excusable. But that is not the point. Sense and coherency is.

Without psychology, characters are assembled from parts. The brooding loner. The power-hungry tyrant. The cold manipulator. You’ve seen these figures a thousand times because writers keep grabbing the same pieces off the shelf and stitching them together. The result is a character that functions, technically, but never surprises. Never feels like a person who might exist off-page. Those can haunt.

With psychology, characters become inevitable. The reader can trace the forces that made them. They recoil from the outcome but they understand the machinery. They can’t dismiss the villain as simply evil, because they’ve seen the path. They know, in some uncomfortable way, that the path was walkable. That anyone could have walked it, given the right wounds and the wrong circumstances.

The gap shows most clearly in antagonists. A psychologically literate writer can articulate why their villain believes they’re correct. An illiterate one just makes them cruel and calls it characterisation.

Why This Is Hard or: The Ego Problem

Here’s a truth: most writers only create characters they’d want to be friends with.

Protagonists get the author’s best qualities, or the qualities the author wishes they had. They’re brave when it counts. Kind beneath the rough exterior. Misunderstood but ultimately good. The protagonist is a wish-fulfillment proxy, the author’s idealised self moving through a world that will eventually recognise their worth. And even the most “anti-hero” is really just a soft, fuzzy thing that oh-so-secretly wants cuddles and love.

Antagonists get the opposite treatment. They become receptacles. Everything the author fears, despises, or refuses to examine in themselves gets poured into the villain. The result is a figure that exists only to be Other. Easy to hate. Morally uncomplicated. Safely distant from anything the author might have to own.

This is projection wearing a plot.

The problem isn’t that it’s lazy—though it is. The problem is what it reveals. When your villain is cardboard, you’re telling the reader something about yourself. You’re saying: I have never genuinely inhabited a worldview I find threatening. I’ve never asked what it would take to make me into someone I despise. I don’t understand people who disagree with me, and I’ve decided that’s their failure, not mine.

That’s not really characterisation, is it.

And I know that if you’ve read this far you’re the likely sort to agree and so, perhaps, I’m preaching to the choir when I say this but: Your work will be so much more impactful if you don’t merely use your own ideology to build characters around but rather use it to inform your exploration of another.

Or your cast will be inhabited by a dozen flavours of you. A dozen flavours of, perhaps, your most palatable beliefs.

The ego wants safety. It wants to write heroes who validate your self-image and villains who confirm your moral superiority. It resists the method acting work because that work is threatening. To do it you must admit the villain is in you somewhere. You have to find that seed and water it enough to watch it grow. Most people would rather not know what flowers.

The Formula Problem or: Where Shallow Writing Comes From

This is where formula fantasy fails hardest.

You know the shape. The Dark Lord wants power because Dark Lords do that. The villain is cruel because cruelty is villainy and villainy requires cruelty. There’s no interiority. No sense that this person believes in what they’re doing. No coherent psychology beneath the armor and the speeches about domination or snapping or whatever else.

The villain exists because the structure requires a villain. That’s it. That’s the explanation.

I won’t name specific books, but you already know the ones I mean. The villains who monologue about darkness and power as though those are motivations rather than aesthetics. The antagonists who do evil things for evil reasons, tautologically, because the author never stopped to ask what a real person would want in that position. You’ve read these books. You might have loved them when you were young enough not to notice the scaffolding.

The result reads like fiction written by someone who has never met a human being with genuinely different values. And certainly not from out behind a keyboard while blocking and calling them a bigot or phobe or whatever emotionally satisfies them that they did the right thing not trying to understand them.

Really, people we disagree with aren’t always wrong, but rather are someone who arrived at their conclusions through a coherent process you could follow if you but tried. The villains in formula fiction aren’t people. They’re obstacles. Abstractions to be overcome so the hero can complete their arc.

This is what happens when writers skip the psychological work. When empathy stays performative. When the ego protects itself from the contamination of genuinely understanding the opposition. You get villains who function mechanically but collapse under the slightest scrutiny. Who exist to be defeated rather than understood.

And readers feel it, even when they can’t name it. They finish the book and forget the antagonist’s name by the following week. Nothing lingers. Nothing haunts. The villain was never real enough to leave a mark and no amount of CGI or dramatic prose can make that less true.

The Craft Principle

If you can only write characters you like, you’re not writing fiction. You’re writing propaganda for your ego.

So then the work is to understand people. All of them. Including the ones whose existence makes you uncomfortable, whose beliefs threaten yours, whose choices you find repugnant. Psychology gives you the scaffolding: the attachment styles, the defense mechanisms, the shame and wound and compensation that drive human behavior beneath the surface. Method acting gives you the practice: the discipline of asking what would I do rather than what would a villain do.

The result is characters who feel like they exist independently of your approval. Who breathe on the page because you’ve breathed through them. Whose interiority is so coherent that readers can’t dismiss them, can’t write them off, can’t maintain comfortable distance.

I’ve written before about what this produces on the page.

This essay is about what it requires from the author. The psychological risk. The ego dissolution. The willingness to find the villain inside yourself and understand them well enough to write them true.

The alternative is children’s morality plays. Stories where the good people are good because they’re like you, and the bad people are bad because they’re not. Safe. Predictable. Forgettable.

You can write that if you want. But don’t mistake it for craft.

Next : Why Clair Obscur: Expedition 33’s Game of the Year win matters for worldbuilders, and the uncomfortable question the game asks anyone who builds fictional worlds.

Subscribe now

Fair winds,
D. S. Black

Primal has no dialogue.

None. Two seasons of gorgeously rendered violence, grief, found family, betrayal, sacrifice, love, rage communicated entirely through action, expression, and le non-dit.

This shouldn’t work. Every writing manual insists dialogue is essential. Every screenwriting course teaches you to reveal character through what people say. Prose workshops drill you on subtext within conversation, on the telling pause, on what characters mean versus what they state.

Genndy Tartakovsky ignored all of it. And made one of the most emotionally devastating pieces of fiction in the last decade. If you asked me, that is.

This essay isn’t a review. I’m not sure I’m qualified for that, but I can dissect why I believe its wordless storytelling works, and what prose writers, game designers, and anyone building narrative can take from it.

The Hierarchy

Tartakovsky understand that, for this series, body language is the foundation. Dialogue is furniture. Or rather, there is simply no dialogue and body language might very well be exactly how the earliest of humanity spoke.

We’ve been trained to think of physical description and body language as seasoning—the “he crossed his arms” you sprinkle between lines of speech, the “she looked away” that’s meant to signal emotional subtext. Beats. Or stage direction.

Primal inverts this. The physical behavior isn’t supporting the emotional content but rather is the emotional content. When Spear grieves, we see his body curl inward, see the way he holds space around the absence. When Fang protects, we see her position herself between threat and ally before the threat even climbs into frame. When trust fractures between them, we see the physical distance open—literally. With Spear sitting much further away.

No internal monologue to clarify and make sure we understand. No dialogue to state the subtext. Just bodies in a space.

Demonstration

Theory is cheap. Let me attempt to demonstrate. For fun, if anything.

Here’s the same emotional beat written three ways: a subordinate announces he’s transferring to a new master. The man who “collected” him—who views ownership of people as existential—receives the news. The subordinate, who has always seen more than he let on, chooses a moment to stop extending that courtesy.

Version 1: Dialogue-Forward

“I’m returning this.” Calix held out the bolt pistol. “My new master provides his own tools.”

“Your new master.” Saren let the words hang. “Origen.”

“Yes.”

“And what does he see in you, I wonder? The controlled violence? The useful savagery?” Saren’s laugh was soft, almost admiring. “He’ll catalog you. File you away in that vast archive of his. Is that what you want? To be understood?”

“You speak as if understanding is a threat.”

“For men like us? It is.” Saren was quiet for a moment. “He’ll find the hollow places, Fellner. The ones you’ve papered over. He’ll name them. And once something is named, it can be used.”

“Like you used mine.”

“I gave yours purpose. There’s a difference.”

“Is there?” Calix’s voice was quiet. “The storm you hold back must be immense.”

Silence. When Saren spoke again, something had changed. “Origen taught you that. That particular cruelty.”

“No. I’ve seen it since Margard. I simply never had the words.”

This… works (but feels damn strange to me.) The subtext is present, I suppose. Saren’s fear of being known, his possessiveness framed as protection, Calix’s deliberate withdrawal of a courtesy he’d been extending all along. A playwright could stage this. Pinter could probably make it sing.

Version 2: Dialogue + Integrated Beats

“I’m returning this.” Calix held out the bolt pistol, grip-first. “My new master provides his own tools.”

Saren didn’t take it. He circled instead, boots tapping a slow rhythm on the deck. “Your new master. Origen.” He stopped at the viewport, silhouette framed against the burning star. “And what does the old scholar see in you, I wonder?”

“Purpose.”

“Purpose.” Saren turned, smile playing at his mouth. “He’ll catalog you, Fellner. Every trauma indexed. Every wound cross-referenced.” He moved closer. “Is that what you want? To be understood?”

Calix remained still, the pistol extended, untaken.

“For men like us, understanding is violation.” Saren’s voice dropped. “I never named your hollow places. Did you notice? I let you keep them.”

“And I extended you the same courtesy.” Calix met his gaze. “Until now.”

Saren went still.

“The storm you hold back must be immense.”

The smile didn’t fade so much as freeze in place. Something behind Saren’s eyes shuttered.

“Origen taught you that,” he said finally. “That particular knife.”

“No. I’ve seen it since Margard. I simply never chose to say it aloud.”

Also competent. The movement creates tension, the beats add information rather than merely labelling emotion, the physicality supports. This is prose fiction—physical behavior integrated with speech, each informing the other. Most published work lives in this register.

Version 3: Body Language Primarily

“A fine weapon,” Saren murmured as he came to a halt before Calix, his mismatched eyes not on the pistol, but on Calix’s face. “A tool for a precise hand. It suits you. Keep it.”

“My new master provides his own tools, Lord Captain,” Calix replied.

The rhythmic cadenza of clicking bronze heels ceased. Saren went utterly still, his head tilting fractionally, like a predator that has just caught an unexpected, dangerous scent on the wind.

“Your... master,” Saren repeated, the words a low purr.

He reached out, his movement a meander rather than an advance. His gloved fingers, splayed, did not touch Calix’s face, but instead went to the high collar of his new uniform coat. He adjusted it—a paternal gesture that instead felt proprietary.

“Look at you. Dressed in his sober colors, reciting his cold logic.” His other hand settled on the Grave-Pelt at Calix’s throat, thumb brushing against dead bone. “But beneath it all... the beast I found still lingers.”

For a long moment, the focus in his eyes wavered. The performance seemed to pause—not into anger, but into something older. The gaze, ordinarily sharp, softened, becoming distant. Lost. It was the look of a boy adrift in a cold and empty void.

Calix saw it. Searched the face with a saccadic eye.

“The storm you hold back must be immense.”

Perhaps a bit overwrought here, with more telling than showing, but you get the point.

What Each Version Can Do

These aren’t failures and successes, per se. Every writer often renders their chosen technique brilliantly, if they care to become competent at it.

Some revelations, spoken aloud, become smaller. Some overwrought physical description could instead be a killer bit of dialogue. And sometimes, marrying them both is the answer.

The moment Saren says “you’ve hurt me” or Calix says “I see through you,” the power drains from the scene.

This isn’t prescriptive. Dialogue-forward writers exist and thrive—Mamet, Pinter, Elmore Leonard. Their work does things mine certaintly cannot. The stage has constraints that demand speech carry weight prose can distribute elsewhere.

The question isn’t which approach is correct. It’s whether you’re making the choice consciously through your voice.

But more to the point

Tartakovsky isn’t working in prose. He’s working in pure visual sequence—animation that will never have the luxury of interiority.

And yet Primal carries psychological complexity that most dialogue-heavy fiction can’t touch. Spear’s grief isn’t explained. His bond with Fang isn’t declared. His capacity for violence and tenderness aren’t reconciled through conversation. They coexist in his body, visible in how he moves and what he protects.

Games face the same constraint. Combat, traversal, environmental storytelling—wordless by necessity. The narrative designer who understands body language as storytelling tool can make a character’s fighting style communicate psychology, their positioning relative to the player speak relationship.

What can the body say that speech would diminish?

Primal isn’t just “animation without dialogue.” It’s a systematic deployment of physical storytelling techniques that writers may not always consciously learn.

While not tips, just a few things might be worth pointing at to further demonstrate.

Watch where Spear and Fang position themselves relative to each other across the series. Early episodes: wary distance, neither willing to expose their flank. As trust builds, they sleep closer. After betrayal or conflict, the gap reopens.

Tartakovsky never cuts to a character thinking “I’m not sure I trust her yet.” The three feet of empty ground between them says it. When that distance finally closes and Spear sleeps against Fang’s side for the first time—the audience feels the magnitude precisely because no one announced it. It’s just earned.

Establish the baseline so the deviation can speak.

The Test

Take a scene you’ve written. Strip out all dialogue.

Does the scene still communicate? Can you follow the emotional arc through pure physical behavior—who moves toward, who retreats, who can’t meet eyes, whose hands betray what their words hid?

If yes, your dialogue is doing its proper job. That is punctuation, emphasis, the precise word at the precise moment. The structure is beneath.

If no—The dialogue isn’t enhancing; it’s compensating. And your scene will not land with the force it could.

This isn’t about removing dialogue from your work. It’s about building the foundation first. Then adding the words that need to be there—and only those.

Because foundation is built on character psychology and no matter what good actors we are, our words may never be as honest as our actions. Characters should act honestly first.

Primal isn’t a show that happens to lack dialogue. That much is obvious.

It’s a deliberate thematic throughline. And whether the result was purposefully serving this purpose—the ideas that essential emotion can be communicated through action and image alone—it achieved it.

Let someone have their own interpretation of the scene. This anxiety to be sure you’re communicating to the reader/viewer what you need them to understand cheapens the emotional core. And I’m talking about things like explanation that preempts the reader’s own understanding. The interior monologue that hand-holds through subtext anyone paying attention already caught.

Let them take away what they want, not what you believe they should want to take from it.

For more craft analysis, character breakdowns, and worldbuilding deep-dives, subscribe.

Subscribe now

Fair winds,
—D.S.

It’s popular. It circles the drain of overdone.

Jekyll and Hyde. Werewolves. The gentleman who becomes a monster when provoked. The seventeen year old’s deviantArt character with heterochromia.

We’ve been telling this story for centuries. Let’s think about why it’s so fascinating.

I posit: the transformation isn’t the point. The containment is.

Dual-nature characters aren’t compelling because they transform. Althought, for some, I suspect that is absolutely the main draw because it can be all spectacle on television. However, I believe they’re compelling because they contain both states simultaneously—and the tension of that containment is what creates character magnetism.

Somerset—if you permit me my own creation a spot—in his turquoise dress whites, speaking in measured slightly-aristocratic tones, knowing he contains something feral that the sea recognises. Conan the barbarian who understands statecraft better than the kings he deposes. The surgeon whose hands know violence and healing with equal intimacy. The hyena king whose gutter philosophy cuts deeper than any lion’s court rhetoric.

This is one of those “aesthetic obsessions” I never shut up about. Not the cliché split personality, but the more sophisticated construction: characters who are wholly and authentically both things at once. The officer who is also the beast. The scavenger who is the philosopher. The barbarian who is also the statesman.

The Transformation Trap

Most writers treat dual nature as a binary switch. Calm state. Trigger event. Beast mode. Return to calm.

Bruce Banner gets angry, becomes Hulk, smashes things, reverts. The werewolf transforms at the full moon, loses control, wakes up confused. The berserker enters rage, blacks out, surveys the carnage afterward.

I wouldn’t say this model is wrong, per se. I would say it’s a bit flatter than it could be.

It’s circumstantial. The character “becomes” a beast when angry, scared, or lunar-aligned. The duality is something that happens to them, not something they are.

It removes agency. The beast “takes over.” The civilised self is a passenger, not pilot. This is less interesting because the character isn’t choosing anything—they’re being hijacked, in a way..

It’s predictable. Readers know the trigger. They know the result. The tension becomes mechanical: will he get angry? Yes. Will he transform? Oh, yeah. Will he feel bad afterward? Of course.

The best versions of these characters, such as Hulk in recent portrayals where Banner and Hulk negotiate, integrate, coexist, have moved toward what actually works: simultaneous containment rather than sequential transformation.

The switch isn’t the story. The cage is.

Simultaneous Containment

And here is where I reframe the apparatus:

The sophisticated dual-nature character doesn’t become the beast when provoked. They are always both, and they choose which face to show.

The civilised exteriour doesn’t so much suppress the primal as it silhouettes it when it emerges. It displays through contrast.

Think about what a uniform actually does. Naval dress whites, aristocratic protocol, the measured cadence of command voice—these aren’t hiding the predator underneath. They’re framing it. The cage makes the beast legible. Without the bars, you can’t see what’s pacing inside.

In my manuscript there’s a moment where Captain Somerset takes the wheel during a storm that should kill everyone aboard. Here’s what his first lieutenant sees:

His eyes were wide with focus. His shoulders strained against the fine turquoise wool of his uniform coat, the elegant white countershading along the inner sleeves and flanks were a blinding flash against the bruised and black sky. He looked like a predator. He looked like prey. He looked like a man the sea had already claimed but who refused to acknowledge it.

Predator and prey. Officer and beast. In the same sentence. The uniform doesn’t hide what he is—the straining wool, the countershading designed to echo sacred dolphins, the formal dress barely containing something feral. It reveals it through the tension.

This is the principle: not transformation, but containment. Not “he becomes dangerous when pushed” but “he is always dangerous, and what you’re seeing right now is how he holds it.”

Three Expressions: Somerset, Gore, and Daud

The dual nature doesn’t have one shape. In my cast, three characters demonstrate three different expressions of the same principle.

Somerset’s charm is a lie. His rakish (oh, I love that word) persona coalesced in a moment of humiliation. And that moment created the persona: seductive, dangerous, magnetic. All to contain something his culture really doesn’t want to put a name on.

The sea recognises what’s underneath. In his world, the sea is the ultimate predator. Sentient, predatory and starving. To be seen by the hungry ocean is to have your capacity for violence recognised and to therefore open a connection to it. To read it and be welcomed in to drown. Or to keep you and your ship out of reach and stay afloat through listening to the song it sings.

But to speak that language of the sea—the one of something vast and hungry—you need be be just like it.

The gentleman is real. The beast is real. The tension between them is what makes him a witch-captain.

Then there’s Lieutenant Gore. His dual nature inverts the expectation. Where Somerset’s beast is passion barely contained, Gore’s is coldness.

When a Navigator goes missing—slipped overboard and lost—Gore delivers the news:

He said it as if he were explaining a mechanical failure. A component that had exceeded its tolerances and failed. Unfortunate, but predictable.

“Already done, sir.” He tapped the ledger under his arm.

Death logged, filed, processed. His beast—the mirror-image to his containement—is the reptilian efficiency that can catalog a soul and move to the next task without so much as pause.

His aristocratic rulebook, his obsessive adherence to regulation, his ice-cold social formality—these aren’t suppressing his emotions. They’re the shape his predation takes. The civilisation is the weapon.

Some readers expect “cold” characters to secretly have warmth underneath. Gore doesn’t. His inability to feel warmth isn’t a flaw to overcome. Short of making him interesting on the page, it’s the feature that makes him devastating. It’s the point.

So, Daud van Richter, the operative who becomes Somerset’s unlikely mirror—And I keep telling you, I swear it’s not a romance—demonstrates a third variation: beast as profession.

Daud’s knife found the space between the fourth and fifth rib. It was not a dramatic thrust. It was not even particularly violent a motion.

And afterward. No catharsis:

He washed his hands, watching the faint pink swirl away (…) adjusted his coat, smoothing the severe lines of the Befruoren cut.

Mind already on the next variable. The containment continues after violence. That’s the beast as profession.

No berserker rage. No loss of control. No transformation. Daud is a killer the way a surgeon is a surgeon. Whether by training, practice, or craftsmanship. His “civilised” presentation (the elegant fingers, the measured voice, the patience) isn’t containing something wild. It’s containing something merely professional.

The principle across all three, if I’m to sound like I know anything about what I’m talking about and not merely winging this: the “primal” doesn’t have to mean rage. Somerset’s beast is passion. Gore’s beast is precision. Daud’s beast is professional competence at killing. The containment is what matters, not the specific shape of what’s contained.

What matters is that there are two things simultaneously true about one character that, when pulled apart, seem incompatible.

The Inversion Principle: Conan and Chaa

The dual nature works in either direction. The interesting characters aren’t always officers containing beasts. Sometimes they’re beasts containing officers.

Robert E. Howard’s Conan—whom I adore to pieces—isn’t compelling because he’s a barbarian. He’s compelling because he understands statecraft, reads political situations with a general’s eye, and recognizes civilisation for what it is—Which is organized savagery with etiquette. And through his ability to inhabit both worlds, is able to offer precise observations that make you consider about how absurd it all is.

“Barbarism is the natural state of mankind. Civilisation is unnatural. It is the whim of circumstance. And barbarism must ultimately triumph.”

The primal exteriour contains strategic intelligence. Conan survives throne rooms and battlefields because he operates in both of these registers. The barbarian contains the king.

One more. Chaa, my hyena king from Clawstar—a different project, genre but same bones—takes that further. Chaa is a literal scavenger. He’s a second-tier male in a matriarchal clan, destined for subservience and scraps. When the lions’ “righteous cull” murdered his queens, his sisters, his matriarchs, he filled the power vacuum not with nobility or tradition but with teeth and cynical clarity.

“Their law is the cage, and they are surprised when the prisoners rattle the bars. Let the dark come. At least it is honest.”

Cynical wisdom forged in the gutter from of the mouth of a low-life. The prides who cull his kind believe they’re civilised. Chaa understands that their civilisation is just violence with cleaner aesthetics.

To his followers, Chaa is the necessary monster. To his enemies, he is the embodiment of chaos—never realizing they were the ones who created him.

Chaa isn’t like Conan because he is not romantic. No nobility in his savagery. He’s a king of scraps whose worldview was proven correct by the very powers who look down on him. His wisdom isn’t despite his circumstances. He’s not rising above anything. And his low vantage allows him to see what the apex predator cannot.

Building Dual Nature from the Ground Up

How to do it?

Make it more than circumstantial.

Don’t give them a trigger. Build the duality into their baseline psychology. The question isn’t “when do they become dangerous” but “what keeps them from being dangerous all of the time.”

The cage is visible. Right there.

Uniform. Protocol. Ritual. Manners. Code. Whatever your character uses to structure themselves—it shouldn’t hide the beast. It’s framing the beast. Readers need to see the cage and what’s pacing inside.

Then, let the civilised and primal serve different functions.

The officer makes you effective. Things like strategy, command, social navigation, long-term thinking. The beast makes you dangerous. Survival, violence, instinct, immediate action. Both necessary. Neither is “the real them.”

No “normal mode” versus “beast mode.” The character should be readable as both in every scene. Readers should always be slightly uncertain which face they’re seeing, because both faces are always present. The charm that might be genuine or might be calculation. Keep both possibilities alive.

Make the containment costly.

The cage takes energy to maintain. Protocol is often exhausting. Performance is labour. Let readers see what it costs to hold the beast—whether by drinking, isolation, relationships that can’t survive proximity to something that controlled. The containment shouldn’t always be effortless. It should be the character’s primary ongoing work.

Why This Works in Grimdark

This concept, I believe, has a special place in the thematic throughline of grimdark. Which is why I wanted to mention this on its own.

Grimdark demands characters who can survive horror without breaking. Cozy fiction can have protagonists who are purely civilised because their worlds don’t require predation. Grimdark worlds do.

The mathematics are simple:

Pure civilisation breaks under pressure. It can’t do what survival requires. When violence is necessary, the purely civilised character hesitates, compromises, or shatters.

Pure beast can’t navigate complexity. No strategy, patience, or social intelligence. Raw predation without containment burns out fast—killed by something smarter, betrayed by a need for patience.

Both simultaneously? Devastating.

And the tension between two natures that seem—on the surface—diametrically opposed, is what makes a character unforgettable.

If this resonated, you might also want to read the companion piece on why compelling beats likeable every time:

For more craft analysis, character breakdowns, and worldbuilding deep-dives, subscribe.

Subscribe now

Fair winds,
—D.S.

The protagonist of my novel The Reply is not a “good” person. Certainly not in the modern definition. What he is: perfectly adapted to his world.

Captain Henry Somerset is charming—but it’s performance, a weapon he wields to disarm and seduce. He treats women like conquest trophies, drinks too much, and generally channels his considerable trauma into becoming excellent at violence. His loyalty is fierce but possessive. His competence borders on inhuman. When he smiles, it’s calculation, not often warmth.

He’s also the most compelling character I’ve ever written.

In Nhera, where the ocean is sentient and predatory, where competence is the only thing standing between you and drowning, Somerset is exactly what survival requires. He’s as dangerous as the world that forged him.

And yet every writing workshop, every social media thread, every virtue-signaling checklist would tell me he’s “problematic.” That I should soften him, redeem him, make him learn to be kinder.

Because here’s what modern writing advice gets catastrophically wrong: sympathy is not the same as compelling. And the relentless push to make protagonists “likeable” is producing fiction that’s predictable, safe, and—worst of all—boring.

The Sympathy Trap

We’re told protagonists must be:

• Kind (or trying to be)

• Morally legible

• Motivated by care for others

• Redeemable through growth

• Fundamentally good

This isn’t craft advice so much as ideology masquerading as technique.

The moving target of what counts as acceptable character behavior shifts with political winds. What’s “sympathetic” in 2025 would’ve been unrecognizable in 2015. Writers tie themselves in knots trying to hit a standard that changes faster than they can revise.

The result? Protagonists who are safe. Predictable. Designed by committee to offend no one.

If you always know your protagonist will choose compassion, help the vulnerable, and learn to be better—what’s the point of reading? That’s not narrative tension. That’s cozy political porn.

The Diegetic Problem: When the Author Shows Their Hand

Here’s the craft issue: authorial judgment kills immersion.

When your narrative voice signals disapproval of a character’s choices—when the prose itself leans in to let readers know “this is bad and you should feel bad about it”—you’ve broken the fictional dream. You’re no longer in the story. You’re being lectured about the story by someone who needs you to have the correct opinion.

Diegetic writing—fiction that stays inside the world without external commentary—requires neutrality. Not moral relativism. You present the character’s logic, their context, their choices, without the narrative voice editorializing.

Example of non-diegetic writing:

Somerset smiled that cruel, predatory smile that revealed everything ugly about his treatment of women, his need to dominate, his fundamental brokenness that he refused to address.

Diegetic version:

Somerset smiled.

The first version is the author controlling your interpretation. The second trusts you to see what’s happening and form your own judgment. One is propaganda. The other is fiction.

Fiction isn’t an instruction manual. It’s not modeling correct behavior. It’s exploring what humans do under pressure.

When you write morally complex characters without narrative judgment, readers engage authentically. They think. They debate. They feel complicated things about people doing complicated things in complicated circumstances.

The moment you signal which opinion you want them to have, you’ve turned fiction into a morality play. And readers who came for story, not sermon, check out.

There’s disposable fiction—stories consumed and forgotten. And then there’s fiction that stays with readers for years because the author trusted them to form their own interpretation. When you let readers build their own relationship with the text, when you resist the urge to guide them toward the “correct” takeaway, you create space for genuine engagement.

Writing that aspires to educate readers on morality is a virtue signal, not a snapshot of human experience. And virtue signals don’t stick with anyone—they just demonstrate the author performed the right-opinions-at-time-of-publication.

Why Unlikeable Protagonists Work: The Somerset Case Study

Let me be specific about why my protagonist works despite (because of?) violating every “likeable protagonist” checklist:

He’s adapted to his environment. Nhera isn’t a world where kindness is rewarded. The ocean is sentient, predatory, and wants you. Ships disappear. Sailors drown. The sea whispers promises and threats in equal measure. In that context, Somerset’s weaponized charm, his possessive loyalty, his refusal to be vulnerable—these aren’t character flaws. They’re survival traits. The world made him dangerous because anything less gets claimed by the depths.

He’s a commoner who clawed his way to Post-Captain through merit alone. The aristocracy despises him for it. His response? Be a jerk, seduce their daughters, drink their wine, and beat them at their own games while smiling like he was born to it.

The charm is trauma response. At his first high-society gala, a noblewoman treated him like an exotic pet—praising his “raw talent” and “unrefined energy” with amused condescension. That humiliation forged his rakish persona. He treats women of that class as conquest to reclaim the power stripped from him. It’s pathological. It’s ugly. It’s rather psychologically coherent.

His competence is the point. Somerset survives because he’s the best naval officer in Arune. Not the kindest one or most moral. His skill at reading storms, navigating impossible waters, and commanding a ship borders on supernatural. The sea itself is obsessed with him.

Readers don’t like him. They’re fascinated by him.

Competence > Sympathy in Grimdark

Here’s what grimdark understands that cozy fiction doesn’t: interesting beats likeable every time.

Somerset doesn’t need to be sympathetic because he’s:

• Complex: His flaws have clear psychological origins. You understand why he’s like this even if you don’t approve.

• Competent: When he takes the wheel in a storm, his crew watches a man become a god. That’s more compelling than any amount of emotional availability.

• Consistent: He doesn’t apologize for what he is. No redemption arc where he learns to be nicer. He’s a weapon pointed at the ocean, and the ocean wants him back.

The moment you make Somerset “sympathetic,” you lose what makes him work. If he starts treating women better, stops drinking, learns healthy emotional expression—he becomes predictable. And predictable characters are narrative dead weight.

Why This Works Across My Cast

This isn’t just Somerset. My entire main cast operates on “compelling > sympathetic”:

Lieutenant Gore: Aristocratic, protocol-obsessed, cold. Loyal to Somerset not from affection but from pragmatic respect for competence. Possibly gay, definitely repressed. Would execute a crew member for insubordination without hesitation.

Daud van Richter: Richter’s bastard half-brother, her deniable knife. Missing molars from a job gone wrong. Kills efficiently, questions rarely. When Somerset forces him to choose between completing his mission or saving Somerset’s life, he chooses Somerset—not from friendship, but from recognition. Two weapons acknowledging each other.

What “Flaws” Actually Mean

Modern writing workshops treat character flaws like:

• Small, manageable quirks

• Opportunities for growth

• Things to be overcome by Act III

That’s not a flaw. That is a plot device with a redemption timer.

Real flaws—the kind that make characters jump off the page—are:

• Structural to their psychology: Somerset’s performative charm isn’t merely bad habit he can unlearn. It’s load-bearing architecture holding up a psyche built on class resentment and childhood humiliation.

• Incompatible with easy redemption: You can’t “fix” Gore’s aristocratic coldness without fundamentally destroying who he is. His inability to feel warmth isn’t a bug—it’s the feature that makes him work as an intelligence officer.

• Morally uncomfortable: Daud’s competence at violence isn’t softened by reluctance or regret. He’s good at killing and knows it. Readers can be uncomfortable with that. Good.

The Permission You Need

If you’re writing grimdark, horror, psychological thrillers, or any genre where stakes are survival rather than personal growth:

Stop trying to make readers like your protagonist.

Make them:

• Competent at something that matters

• Psychologically coherent (even if ugly)

• Consistent in their damage

• Adapted to the world they inhabit

• Interesting enough that readers have to keep reading or watching.

Somerset isn’t sympathetic by Bluesky standards. He’s a traumatized weapon who treats the sea like an abusive lover and his crew like the only family he’ll allow himself. He uses people. He performs constantly. He’s probably going to die badly.

But I can’t stop writing him.

And readers who claim they want “likeable protagonists” keep telling me they can’t stop reading about him either.

There’s no shame in finding complex, dangerous, morally ambiguous characters compelling. This is not a failure of your values. That’s proof you understand that fiction isn’t bloody-well a morality exam.

The shame comes from people who need you to perform the correct opinion about fictional characters—as if your engagement with the Somersets of fiction says something damning about your real-world ethics.

It doesn’t. Fiction is where we explore what we may not tolerate in reality. That’s the whole point.

The Craft Principle

Compelling characters operate on internal logic, not external approval.

Somerset’s psychology makes sense to him. His actions follow from his trauma, his competence, his relationship with the sea. He doesn’t break character to be more palatable. He doesn’t soften for audience comfort.

That internal coherence—that refusal to apologize for what he is—creates the magnetism that sympathy never could.

Sympathy is asking for permission. Complexity is a territorial claim.

When you write for sympathy, you’re asking: “Is this okay? Will readers accept this?”

When you write for complexity, you’re claiming: “This is what this person is. Engage with it or don’t.”

One creates cozy political porn. The other creates crap like The Reply.

I wrote a companion piece about antagonists.

Subscribe now

For more craft analysis and worldbuilding breakdowns, subscribe.

Fair winds,
—D. S. Black

When I tell people I’m building languages for my fantasy world, I usually get one of two reactions:

1. “Oh, like Tolkien!” (Yes, but also no.)

2. “Isn’t that just... making up words?” (Yes, but also very much no.)

Here it is: anyone can smash random syllables together and call it a language. Most fantasy writers do. And most of the time, it sounds like someone sneezed on a keyboard.

Kh’zarthyx’ul. Ae’tharion. Zyx’kael.

You’ve seen it. I’ve seen it. We’ve all seen it. And we’ve all quietly cringed.

But good conlanging—the kind that makes a world feel real—isn’t about sounding exotic. It’s about sounding inevitable. Like these words have been spoken by real people for hundreds of years, worn smooth by use, shaped by the needs of the culture that speaks them.

So how do you do that?

Here’s what I’ve learned from building two languages for my world (Arunaic and Low Aelhir), informed by a lifetime of being a bilingual, bidialectal weirdo who accidentally became a conlanger.

Step 1: Start With Culture, Not Sounds

Most people start conlanging by picking “cool sounds” and mashing them together. That’s backwards.

Start with: Who are these people? What do they need to say?

Example: Arunaic (The Language of Sailors)

The Aruneans are a maritime culture. Their entire civilization is built on ships, trade, and naval power. So their language reflects that:

• 6+ words for wind (shao = breeze, shaul = gale, shaullue = wind caught in sails)

• Depth/distance is EVERYTHING (linne = shallows, laae = deep, drau = abyss)

• Time is measured by the sun’s passage (fenilasra = high passage/noon, feilasra = waking passage/morning)

The vocabulary tells you about the culture. Aruneans don’t just have “one word for ocean”—they have words for coastal waters, deep sea, drowning depths, and the horizon. Because those distinctions matter to them.

Even their color words are depth-based. They don’t see “blue”—they see where in the water column that blue exists:

• muirrine = sea-blue (the color of shallow or near-surface seas)

• laagerrine = deep loden green (the color of the mesopelagic zone)

• nadirrine = abyssal purple-black (the color of crush-depth)

When an Arunean describes something as muirrine, they’re not just saying it’s blue. They’re saying it has the quality of the sea itself—open, deep, unknowable.

But here’s where it gets interesting:

Aruneans don’t just have “a word for travel.” They have laaonarre.

Etymology:

• laae (deep, beyond the coast) + on (across, beyond) + maare (horizon)

• Meaning: Traveling beyond the coast and across the horizon—into the unknown.

What this tells you about Arunean culture:

To an Arunean, real travel isn’t just “going somewhere.” It’s leaving safety behind. It’s crossing into the deep (laae), beyond sight of land, where the horizon (maare) becomes your only guide.

There’s no single English word for this. “Voyage” is close, but it doesn’t carry the weight of risk, of leaving the known world. “Journey” is too generic. Laaonarre is specific. It’s sacred. It’s what separates a sailor (muirar) from someone who just owns a boat.

This is what good conlanging does. A single word reveals an entire philosophy. It shows you what a culture VALUES—and what they FEAR.

Even their military ranks encode this relationship to the sea:

• Maarendar = Captain (”horizon-commander” - master of the reach)

• Draumeir = Admiral (”abyss-master” - lord of crushing deep)

• Fendraumeir = Fleet Admiral (”high-abyss-master” - master of all depths)

Rank isn’t just hierarchy. It’s how deep you’re trusted to sail, how far from shore your authority extends. A captain commands the horizon, but an admiral commands the abyss itself.

Your language should do the same. If your culture is desert nomads, what’s THEIR word for the moment you leave the last oasis and head into open sand? What do mountain-dwellers call the act of descending into the lowlands? What do your characters call the thing they do that NO OTHER CULTURE has a word for?

That’s where language becomes world-building.

Culture shapes language. Always.

Step 2: Choose Sounds That Fit the Vibe

Once you know WHO is speaking, figure out what they should SOUND like.

Arunaic is:

• Vowel-heavy (a, e, i, o, u dominate)

• Flowing, liquid consonants (l, r, n, m)

• Few harsh stops (no hard K or T clusters)

Why? Because it’s a language designed for speaking on ships—over wind, over waves, over distance. You need CARRYING sounds. Long vowels. Resonant consonants.

Compare that to Low Aelhir (my elven language):

• Sharper consonants (kh, th, zh, hard R)

• More guttural (especially in the Draihir dialect)

• Shorter vowels

Why? Because elves in my world are older, harsher, more warlike. Their language reflects that—it’s harder, more angular, less forgiving.

The sound should match the culture.

Step 3: Build Derivation Rules (So You’re Not Just Making Shit Up)

This is where most conlangers fail.

They make up 50 random words, slap them in a glossary, and call it done. But then when they need a NEW word (which will happen constantly), they just... make up another random word. No consistency. No internal logic.

Good conlangs have RULES.

Example: Arunaic Compound Words

Arunaic builds new words by COMBINING root words:

• drau (abyss) + hessa (horse) = drauhessa (drown-horse, a mythological sea creature)

• shaul (gale) + lue (caught/captured) = shaullue (wind in the sails)

• thea (return) + lua (light/beacon) = Thealua (the Return-Light, the great lighthouse of Theastone)

This means I can generate NEW words whenever I need them. I’m not making shit up—I’m DERIVING words from the system I already built.

This compounds beautifully. Once you have hessa (horse) and drauhessa (drown-horse), you can build:

• hessar = rider, horseman

• drauhessir = of/relating to drown-horse heraldry

• allahessen = horse dressage, martial performance (from allan = graceful form + hessa)

Or take something like bibilausa—a word that combines bibi (cute, small, harmless) + lausa (beast, prey). It means “useless but endearing,” the kind of creature that’s too cute to hunt. It’s the Arunean word for a lapdog. One compound tells you that Aruneans view most animals through the lens of utility, and anything that fails that test is... well, adorably pointless.

Your conlang needs this. Otherwise, it’s just a list of random nouns.

Step 4: Make It Speakable (Or It’s Just Decoration)

Here’s a test: Can you say your fantasy words out loud without sounding like you’re gargling gravel?

If the answer is no, you’ve failed.

Bad fantasy names:

• Kh’zarthyx (how do you even pronounce this?)

• Ae’thalos’kyr (three syllables? four? who knows?)

• Xyl’gothrim (unpronounceable)

These aren’t WORDS. They’re PUNCTUATION.

Good fantasy names:

• Muirrine (myoor-EEN)

• Drauhessa (DROW-hess-ah)

• Thealua (THAY-ah-loo-ah)

You can SAY these. They have rhythm. They have flow.

If your readers can’t pronounce your words, they’ll skip over them. And if they’re skipping over your words, your world-building has failed.

Step 5: Let Your Own Linguistic Background Inform Your Work

Here’s my secret weapon: I grew up bilingual and bidialectal.

• Native Spanish speaker

• Native Midwestern American English speaker (from family)

• Native N/W London English speaker (from childhood friends)

This means I’ve spent my entire life code-switching—flipping between languages and accents depending on context. I can HEAR how languages work. I can FEEL when a sound pattern is wrong.

This is why I can build conlangs that feel real.

I’m not guessing. I’m drawing on a lifetime of linguistic immersion.

You don’t need to be bilingual to conlang well. But you DO need to:

• Listen to how real languages sound (not just English)

• Pay attention to rhythm, stress, intonation

• Study how languages EVOLVE (why do some sounds change? why do dialects diverge?)

The more you understand about real languages, the better your fake ones will be.

The Takeaway

Good conlanging is:

1. Culture-first (what do these people need to say?)

2. Sound-appropriate (what should this language sound like?)

3. Rule-based (how do I generate new words consistently?)

4. Speakable (can I actually say this out loud?)

5. Informed by real linguistics (how do real languages work?)

Bad conlanging is:

• Random syllables with apostrophes

• Unpronounceable clusters

• No internal logic

• Just “sounding exotic” for its own sake

If you’re building a fantasy world and you want your languages to feel REAL, start with culture. Build from there. And for the love of all that is holy, make sure your readers can actually PRONOUNCE your words.

Your world-building will thank you.

Want to go deeper? Next time I discuss conlanging, I’ll break down the elven pronoun system I built for Low Aelhir—where ‘you’ and ‘I’ aren’t fixed identities, but shift based on who’s dominating the conversation. It’s a linguistic nightmare. It’s also one of my favorite things I’ve ever built.

If you want to see that (and more craft deep-dives), subscribe. I post every Tuesday.

Subscribe now