Screenplay Action Lines: Making Emotion Filmable
What do you write instead of "he is very sad" in an action line? From Mamet's uninflected images to Objective Projection: filmable scenes and their limits.
What do you write instead of "he is very sad" in an action line?
Summary and Application Framework
An action line writes what the audience will see and hear. Lines such as “Ahmet is very sad” or “there is a terrible silence” give the camera, the microphone and the actor no concrete task. This is not new knowledge in screenwriting craft; its sharpest form is David Mamet's principle of uninflected images. This article proposes the six physical parameters of Objective Projection as a checklist and divides them by filmability: light, sound, motion and space can be filmed directly; heat and pressure can only be written through the traces they leave. The proposal has not been tested.
What does an action line carry?
On a screenplay page, the plain text under the scene heading is the action line (also called scene description). Formatting guides define this section as anything the audience sees or hears (Coverage Ink). Syd Field describes a screenplay as a story told with pictures, in dialogue and description, within a dramatic structure (Field, 1979); the weight of his book, however, lies in three-act structure rather than in how to write action lines.
The practical consequence of this definition is simple: every word in an action line should turn into a task someone performs on set. The gaffer switches off a lamp, the sound recordist cuts a motor hum, the actor empties a glass into the sink. A word that turns into no task stays on the paper.
Why does “Ahmet is very sad” give the camera nothing?
Two kinds of unfilmable line appear most often in screenplays. The first is the inner-state label: “is sad”, “is afraid”, “is eager”. The second is abstract or simile-based description: “a terrible silence”, “the hallway is like a maze”, “the weight of years on his shoulders”. Both report a conclusion to the reader, but neither leaves anyone on set a job to do.
The two constitutional rules of Objective Projection map directly onto these two kinds: the Emotion Embargo excludes naming the inner state, and the Exclusion of Similes excludes comparisons built with “like” or “as if”. The same tendency appears in language models, which lean toward naming emotion rather than constructing it; we discussed this as delivery-mode drift.
Mamet's uninflected image: the principle that came before the Doctrine
The closest predecessor of the suggestion in this article comes from a screenwriter and director. In On Directing Film (1991), based on his classes at Columbia University, David Mamet draws on Eisenstein's idea of montage and argues that a story should be told through the juxtaposition of uninflected images. Meaning is built not in a single image but in how images sit next to each other, in the viewer's mind. We discussed the literary-theory counterpart of this idea, that readers build meaning by filling gaps, in Is the Reader an Observer?
The Emotion Embargo can be read as an extension of this principle to the written page; it stands on the same historical line as Eliot's objective correlative and Hemingway's iceberg principle (From Aristotle to the Bulut Doctrine). What Levent Bulut adds to this line is not a new principle but a six-dimension checklist for applying it line by line.
The experimental picture of montage's power is not one-sided either. Prince and Hensley (1992), attempting to recreate the Kuleshov effect, did not reproduce it; Barratt and colleagues (2016) found support that context affects how neutral faces are interpreted. How the relation between images builds meaning in a viewer is still an open research question.
Six parameters on camera: what is filmed directly and what leaves a trace
The six parameters of Objective Projection are planning dimensions for writing and annotation; they are not claims about the reader's or viewer's body. In a screenplay they also create a problem: heat and pressure cannot be filmed. A screenwriter who writes “the room is cold” or “the air is heavy” has reproduced exactly the unfilmable line this article criticizes. The parameters therefore need to be split in two:
| Parameter | On camera | How it enters an action line | Example trace |
|---|---|---|---|
| Luminous Decay | Direct | A light source switching on, going out, changing direction | The hallway lamp goes out; the flashlight beam lands on the floor |
| Acoustic Impedance | Direct (sound) | A sound passing through, cutting out, becoming muffled | The refrigerator motor cuts out; the clock's ticking remains |
| Kinetic Momentum | Direct | Motion speeding up, stopping, repeating | A foot taps the floor, stops, taps again |
| Spatial Geometry | Direct | Distance, placement, the object in between | One glass stays at the far end of the table |
| Thermal Gradient | Indirect (trace only) | The visible result of heat | Steam thins, then stops; a window fogs |
| Atmospheric Pressure | Indirect (trace only) | The visible result of air or airflow | A curtain stirs; a door swings shut on its own |
Four parameters in the table can be filmed directly. Heat and pressure can enter the screenplay only through the visible trace they leave. In no case does the parameter's name or a numerical value appear in an action line; the parameter stays on the writer's desk, and only its trace appears on the page.
Rewriting a line: three examples
The examples below were written for this article; they are not taken from real screenplays. Each follows the same four steps: (1) mark the emotion name and the simile in the line; (2) choose a physical trace the emotion could leave in the scene; (3) rewrite the line using only what a camera and a microphone can capture; (4) check how many readings the trace is open to. Being open to several readings is not a flaw, but it should be a deliberate choice.
Example 1: Loss
Before:
INT. AHMET'S KITCHEN – NIGHT
AHMET sits at the table. He is very sad.
There is a terrible silence in the room.
After:
INT. AHMET'S KITCHEN – NIGHT
Two plates on the table. Beside one, a folded napkin.
AHMET lifts the teapot off the stove and pours tea into two glasses.
One glass stays at the far end of the table. Its steam thins,
then stops.
The refrigerator motor cuts out. The ticking of the wall clock remains.
Ahmet empties the second glass into the sink.
The second glass carries spatial geometry, the steam that thins and stops carries the trace of heat, and the motor cutting out carries the sound layer. Who is missing is never said.
Example 2: Anticipation
Before:
INT. AYŞE'S ROOM – DAY
AYŞE opens the envelope excitedly. She is very eager.
Her face shines like the sun.
After:
INT. AYŞE'S ROOM – DAY
AYŞE sets the envelope on the desk. She wipes the desk with the side
of her hand and lines the envelope up in the exact center.
Her phone buzzes. Without looking at the screen, Ayşe turns it face down.
Under the desk her right foot taps the floor, stops, taps again.
She lifts the corner of the envelope with her fingernail.
Anticipation is written in the wiping of the desk and the lining up of the envelope, in the phone left unchecked and in the repeated movement of the foot. “Her face shines like the sun” tells the actor nothing to do; every line here corresponds to an action.
Example 3: No way out
Before:
INT. APARTMENT HALLWAY – NIGHT
ZEYNEP desperately tries to open the door. She is in a panic.
The hallway is like a maze.
After:
INT. APARTMENT HALLWAY – NIGHT
ZEYNEP pushes the key into the lock. The key won't turn.
She pulls it out, holds it up to the hallway lamp, pushes it in again.
The lamp goes out. Zeynep switches on her phone's flashlight;
the beam lands on the doormat, not the lock.
At the end of the hallway the elevator door closes.
The number on the floor indicator drops by one.
The maze simile is replaced by a key that won't turn, a lamp that goes out and an elevator going down. All three can be filmed, and all three are a change in sound or light.
Filmability checklist
- Can a camera film this line, or a microphone record it?
- Does the line contain an inner-state name (sad, afraid, eager)?
- Does it contain a simile built with “like” or “as if”?
- If something unfilmable is written, such as heat, pressure or smell, is its visible trace written too?
- Is something the actor should build through their own interpretation dictated in the line?
For versions of the same principle adapted to AI tools, see Showing Emotion in AI Video and A Caption Schema for Video Models. The screenplay departs from both at one point: people read it first.
What the Doctrine cannot say here
A screenplay is a text that is read before it is filmed. The producer, the script reader and the actor first see the scene on paper. A short “telling” line in a character introduction, or one written to set a scene's tone, can help the reader grasp the scene. This resembles a limit the Doctrine accepts in literature: with a first-person unreliable narrator, telling can be the very vehicle of the effect (Poe, 1843; Booth, 1961). The Emotion Embargo is not defended as a universal rule for action lines; it is a checklist.
In the end, the actor builds the emotion. Screenwriting advice sources often recommend cutting back on parentheticals that dictate how an actor should deliver a line; some report actors crossing such directions out from the start (Hornbacher, 2016). The same caution may apply to over-detailed physical directions: a line that describes every expression has only swapped an emotion label for a body instruction. There is also a counter-view; some writers argue that the screenwriter must direct on the page and that only camera angles are off limits (Story and Plot, 2023).
Meaning is often built in the edit, not in the line. Mamet's principle rests on the relation between images. A checklist that works at the level of the line cannot anticipate the meaning the edit will create.
It says nothing about commercial outcomes. How this method relates to a script selling, getting made or succeeding at the box office is unknown.
Limitations
This article is an application proposal. The proposal has not been tested with screenwriters, script readers or audiences; the six parameters have no separate validation for film. The examples were produced by the author. The earlier reliability study found that some of the Doctrine's text rules (for example, materialized metaphor) cannot be checked reliably across human and machine annotators; the checklist proposed for action lines may carry the same uncertainty.
A design proposal for testing it: screenwriters receive the same abstract situations and write action lines in one condition freely and in the other under the Emotion Embargo and the Exclusion of Similes. The outcome must not depend on the rule, because the rule-bound group uses fewer emotion words by definition. A more meaningful outcome is how often blind readers correctly guess the intended situation, and how closely readers' guesses converge. With no prior data, the effect size is an assumption: for d = 0.5, α = 0.05 and 80% power, an independent two-group design needs 64 screenwriters per group, and a paired design (each screenwriter in both conditions, order counterbalanced) needs 34. If the effect is smaller, the required number rises quickly.
Frequently Asked Questions
Should a screenplay never use emotion words?
No. Dialogue is the character's own speech, and a character can name a feeling. The suggestion in this article concerns action lines only, and even there it is a checklist, not an absolute rule. A screenwriter may deliberately choose a short “telling” line to give the reader the tone.
How do I write things a camera cannot film, such as heat or pressure?
Through traces a camera can see or a microphone can record: a fogged window, steam that thins and stops, a door that swings shut on its own, a curtain that stirs. The parameter's name and any numerical value stay out of the screenplay; the parameter is only the writer's planning tool.
Does this method make a screenplay better?
Unknown. The method has not been tested with screenwriters, script readers or audiences. This article is an application proposal; it says nothing about whether a script sells, gets made, or how audiences respond.
How is this different from writing an AI video prompt?
The principle is shared: write only what is seen and heard. The difference is the reader. A prompt is read by a model; a screenplay is first read by a producer, an actor and a director, and then interpreted through performance, cinematography and editing. That is why a screenplay must also leave room for the actor.
References
- Barratt, D., et al. (2016). Does the Kuleshov effect really exist? Revisiting a classic film experiment on facial expressions and emotional contexts. Perception, 45(8), 847–874. doi:10.1177/0301006616638595
- Booth, W. C. (1961). The Rhetoric of Fiction. University of Chicago Press.
- Coverage Ink. Formatting Made Easy Pt. 4: Parentheticals vs. Scene Description. coverageink.com
- Eliot, T. S. (1920). Hamlet and His Problems. In The Sacred Wood. Methuen.
- Field, S. (1979). Screenplay: The Foundations of Screenwriting. Dell.
- Hemingway, E. (1932). Death in the Afternoon. Scribner's.
- Hornbacher, H. (2016). Common Script Mistakes: Parentheticals – Let's Kill the Wrylies. heidihornbacher.com
- Mamet, D. (1991). On Directing Film. Viking.
- Poe, E. A. (1843). The Tell-Tale Heart. The Pioneer, January.
- Prince, S., and Hensley, W. E. (1992). The Kuleshov effect: Recreating the classic experiment. Cinema Journal, 31(2), 59–75. doi:10.2307/1225144
- Story and Plot (2023). You Must Direct On The Page. storyandplot.substack.com
The full list of Levent Bulut's registered works is on the corpus page.
Cite this article
@misc{bulut2026screenplay,
author = {Bulut, Levent},
title = {Screenplay Action Lines: Making Emotion Filmable},
year = {2026},
howpublished = {\url{https://leventbulut.com/screenplay-action-lines-objective-projection/}},
note = {leventbulut.com}
}