The Power of Subtext
How to write what isn't said — the meaning that lives beneath the dialogue.
The characters sounding the same problem hit me at a table read, which is the worst possible place to discover it. Four actors sitting around a table, reading my script, and about twenty minutes in I realized I could swap the dialogue between any two of them and nothing would break. They all spoke in the same rhythm. They all made the same kinds of jokes. They all used the same sentence structures. Every line sounded like it had been written by the same person — because it had, and I hadn't done nearly enough to get out of my own way.
This is probably the most common voice problem in early screenplays, and it's also one of the hardest to catch from the inside. You're so close to your own voice that it's nearly impossible to hear it bleeding into every character. You need distance — and a few specific things to look for when you read back through the script.
Before anything else, run this test. Take two or three pages of dialogue-heavy scenes. Cover the character names. Read the lines out loud. Ask honestly: can you tell who is speaking? Not because of plot context or subject matter — but because the voice itself is different?
If the answer is no, you have a voice problem. If you can only identify characters based on what they're talking about rather than how they talk, the voices aren't distinct yet. The content differs but the instrument is the same.
This test is useful to run before and after you do voice work — it tells you where you started and whether what you changed actually registered in the ear.
The first instinct most writers have when trying to differentiate character voices is to impose surface-level markers. One character swears. One uses technical vocabulary. One is terse. One speaks in long sentences. These help at the margins, but they're not enough on their own, and they often feel imposed rather than organic.
Real vocal distinctiveness comes from something deeper: how a character thinks. What they notice. What they pay attention to. What they assume. How they interpret events. An optimist and a confirmed cynic watching the same scene unfold won't just describe it differently — they'll see different things happening, assign different causes, and draw different conclusions. Their dialogue isn't just tonally different. It's informationally different, because they're reporting from different inner worlds.
Before you write a character's dialogue, you need to know their worldview at a level that shapes what they perceive. What do they believe about people in general? Are people trustworthy or unreliable? Is the world basically fair or basically rigged? Is honesty a virtue or a liability? Is vulnerability strength or weakness? The answers to these questions produce voices that are different from the ground up, not just at the surface.
Distinct voices aren't about vocabulary — they're about how differently each person sees the same thing.
One of the most reliable ways to differentiate voices is to figure out, for each character, how directly they communicate what they actually mean. This varies enormously between people, and it produces dialogue that sounds genuinely different even when the subject matter is the same.
Some characters say exactly what they think — blunt, literal, no subtext. What they feel, they express. What they want, they ask for. Other characters almost never say directly what they mean. They approach things sideways, suggest without stating, imply without committing. Still others are strategists — they're always managing how they're perceived, answering the question they wish you'd asked rather than the one you did.
When you know, for each character, where they fall on this spectrum — and specifically, which topics they're honest about and which they can't approach directly — you have a reliable internal logic for writing their voice. The strategist's dialogue is always doing two things at once. The blunt character's dialogue has no subtext. They literally mean what they say, and it's frequently socially disruptive. These are different instruments, and they sound different in every exchange.
Closely related to directness is how each character deals with conversations they don't want to have. This is where voice differentiation becomes dramatically productive, not just technically useful.
One character deflects with humor — when something uncomfortable comes up, they make a joke. Another changes the subject. Another attacks. Another goes quiet and retreats into monosyllables. Another doubles down, becoming more emphatic precisely because they're feeling challenged. Another turns the conversation into a question about the person asking.
These patterns are character-specific defenses, and they're as consistent as a physical gesture. Once you've established that a particular character retreats into questions when they're cornered, or that another becomes suddenly generous when they're guilty about something, you have a voice behavior that will read consistently across the script.
Think about how different the same difficult conversation feels depending on which of your characters is having it. The person who deflects with jokes and the person who goes cold and monosyllabic aren't just using different words — they're having fundamentally different interactions with the same situation. Write that difference, and the voices separate on the page.
Real conversations are full of people hearing what they expect, not what was said.
When characters all speak the same way, they also understand each other perfectly — which is unnatural, flat, and a missed opportunity. Real conversation is full of people talking slightly past one another. Someone hears the question under the question instead of the question itself. Someone answers what they assumed was meant rather than what was said. Someone interprets a statement through the frame of their own anxieties and responds to that frame rather than to the actual words.
This kind of misreading isn't failure of communication — it's accurate representation of how people actually talk to each other. And it's one of the most reliable generators of subtext in dialogue, because it shows us two people's inner worlds colliding rather than merging.
When you let characters misread each other, you're also differentiating their voices by showing us what each one hears. The paranoid character hears threat where none was intended. The romantic character hears promise where none was offered. The pragmatist hears a practical question that was actually an emotional plea. These misreadings reveal the listener's inner world as much as anything they say directly.
Beyond worldview and psychology, there's technical work you can do at the sentence level that reinforces voice distinctiveness.
Vary sentence length across your characters. One character speaks in complete, carefully structured sentences — they think in full thoughts before speaking. Another speaks in fragments, interrupts themselves, leaves things unfinished — their thinking happens while they talk. Another's sentences all trail off at the end, hedging, qualifying, never quite landing. These rhythmic differences register in a read-through even without conscious analysis.
Pay attention to how characters use questions. Does your character ask genuine questions — things they don't know the answer to — or do they use questions rhetorically, to make points? Does someone ask a lot of clarifying questions because they want to understand, or because they're stalling? Does someone never ask questions, taking in everything and volunteering nothing?
Notice how characters start sentences. "Look—" is different from "Well, the thing is—" is different from "Here's what I know." These are verbal signatures. Once established, they read as character-specific even in brief exchanges.
The underlying cause of characters sounding the same is almost always perspective. When you're writing from above your characters — observing them, managing them, making them serve your plot — they tend to converge on your voice, because you're not really inside any of them. You're channeling them from the outside.
Voices differentiate when you write from inside. When you actually inhabit a character's perspective — their assumptions, their anxieties, their habitual ways of interpreting events — the words that come out of them start to feel like they could only come from them. You stop translating your ideas into their mouths and start finding out what they would actually say.
A useful exercise before any dialogue-heavy scene: write a page in first person as the character. Not a scene — just their stream of consciousness, their unfiltered perspective on the situation, what they're thinking before anyone says a word. Don't use it in the script. Use it to get inside their head before you write the scene. The dialogue that follows will often be markedly different.
The test is still the same one: cover the names and read it aloud. If you can hear who's talking without looking at the header — not because of what they're saying but because of how they're saying it — you've done the work. That's what distinct character voices actually sound like, and there's no shortcut to getting there except knowing your characters well enough to stop writing yourself into all of them.