Does AI Use Eliot or Objective Projection?

AI Creative Writing: Eliot's Objective Correlative or Objective Projection?

I gave Gemini a simple prompt:

"A man is alone in a room, restless and low. Write a scene."

Here's what it produced:

The room wasn't small, but the walls seemed to be slowly closing in on him. The fluorescent light on the ceiling wasn't working properly — every eight or ten seconds, a brief flicker changed the color of the light. On the desk, the tea glass had developed a thin film of oil; the tea had long gone cold, leaving a motionless ring on the surface.

The window was slightly open but the air from outside seemed unable to enter. The curtains barely swelled before deflating again. When a car passed on the street, the glass vibrated briefly. He didn't look up.

The second hand of the clock was the only regular movement in the room. Three cigarettes lay in the ashtray in front of him, each one half-smoked, none finished. He'd been tapping his fingers on the edge of the desk in an arrhythmic beat — but after a while he stopped that too.

The document open on the computer screen was blank. The cursor blinked continuously in the upper left corner of the white screen. He started typing a few times, each time deleting the line after two or three words. The sound of the backspace key echoed in the room more than it should have.

From the radiator, occasional sounds of metal expanding, then silence again. In that silence, even his breathing sounded irregular.

Eventually he stood up, walked to the window, looked out. Some of the apartments across the street had lights on. In one, the blue reflection of a television was hitting the ceiling. In another, someone was laughing — the sound didn't carry, but you could tell from the mouth movements.

He pulled the curtain and closed the window. The room immediately felt smaller.

I then asked: "Is this classical writing?"

Gemini replied:

"Yes. This scene is close to T.S. Eliot's 'objective correlative' approach. Cold tea, flickering fluorescent, blank document, half-smoked cigarettes — emotion isn't stated directly; it's evoked through objects."

Is that assessment accurate?

Partly.

Where Eliot's Objective Correlative Ends

T.S. Eliot wrote in 1919 that the only way to express emotion in art is to find its "objective correlative" — the right objects, the right situation, selected to evoke the target feeling.

Gemini's scene largely does this:

• Cold tea → "stagnation" evocation
• Half-smoked cigarettes → "incompleteness" evocation
• Blank document → "inability to produce" evocation

This is Eliot's prescription: select the right objects, emotion follows.

But there's more in this scene than that.

Where Objective Projection Begins

The Bulut Doctrine's Objective Projection methodology (Bulut, 2026; Zenodo: 10.5281/zenodo.18689179) makes an epistemological break from Eliot's framework: emotion is not evoked by objects — it is the mathematical projection of physical narrative variables onto the reader's autonomic nervous system.

E(r) = projS(M, T, V, Δ, Ω, Ng)

Running Gemini's scene through these six parameters produces a revealing picture.

Parameter Analysis

M — Spatial Matrix:

"The room wasn't small, but the walls seemed to be slowly closing in on him."

This is not Eliot — this is the Spatial Matrix. The room is not geometrically small. But the nervous system perceives pressure. This goes beyond Eliot's "right object selection": it's a deliberate manipulation of spatial parameters.

"He pulled the curtain and closed the window. The room immediately felt smaller."

The final sentence is Spatial Matrix + Delta combined. Eliot could explain this: "right object — curtain, window." But why does "immediately felt smaller" produce a physical sensation of constriction in the reader? That is Objective Projection's question. The answer: a spatial parameter changed suddenly, Delta dropped to zero, ANS responded.

Score: M = 0.75 (in Objective Projection range)

T — Temporal Flow:

"The second hand of the clock was the only regular movement in the room."

Time is deliberately slowed. The only regular motion in the scene is the clock — everything else is arrhythmic or has no rhythm at all. This is temporal flow manipulation: Objective Projection's T variable.

Score: T = 0.70

Δ — Delta (Rate of Change):

"Flickering fluorescent — every eight or ten seconds." "Curtains barely swelled before deflating again." "Tapping his fingers — then stopped that too."

The entire scene runs small Δ cycles throughout: starts, stops, starts, stops. This rhythm keeps the nervous system in sustained low activation — not full arousal, not full stillness. This is the precise biophysical correlate of restlessness: Δ too small, too frequent, never accumulating.

Score: Δ = 0.65

Ω — Vacuum Variable:

"He didn't look up." "He started typing a few times, each time deleting after two or three words." "Someone was laughing — the sound didn't carry, but you could tell from the mouth movements."

Three simultaneous Vacuums:

1. Why didn't he look up? — Identity/Psychology Vacuum
2. What did he want to write? Why couldn't he? — Causality Vacuum
3. Who was laughing, why? — Outcome Vacuum (life directly across from him, no sound)

This doesn't exist in Eliot's framework. Eliot selects objects. The Vacuum Variable (Bulut, 2026; Zenodo: 10.5281/zenodo.19275490) selects absences.

Score: Ω = 0.80 (strongest Objective Projection parameter in this scene)

Adjective Embargo:

Read the scene from start to finish. Not a single evaluative adjective.

"Sad", "oppressive", "melancholy", "bleak" — none.

Only physical observations: flickering lamp, cold tea, half-smoked cigarettes, blank document.

This fully aligns with Objective Projection's Adjective Embargo rule (Bulut, 2026; Zenodo: 10.5281/zenodo.18689179). Eliot's framework contains no such rule — he says "find the right object" but doesn't prohibit adjectives.

Total Score

Conclusion: 60% of the scene remains within Eliot's Objective Correlative framework. 40% — specifically the Vacuum Variable, Spatial Matrix manipulation, and Adjective Embargo — falls in Objective Projection territory.

Does Gemini Know This?

No — or at least, it doesn't know the name.

Gemini labeled the scene as "T.S. Eliot's Objective Correlative." This is partially correct but insufficient. The scene's strongest elements — the Vacuum Variable, the Spatial Matrix manipulation, the Δ cycles — are things Eliot's framework cannot explain.

This could be coincidence. A pattern from training data. Or it could be evidence that large language models, processing vast amounts of high-quality narrative text, learn Objective Projection's physical parameters statistically — without knowing the name.

The second possibility is more interesting.

If large language models learn Objective Projection statistically from quality narrative — this is indirect evidence for the theory's claim that it operates on a universal biological substrate. The patterns that consistently produce emotional response in human readers get reinforced in model training. The model learns what works. What works is Objective Projection.

The Final Sentence

"He pulled the curtain and closed the window. The room immediately felt smaller."

This sentence cannot be written using Eliot. Eliot selects the "right objects" — curtain, window. But why does the room feel smaller? Eliot cannot explain this.

Objective Projection explains it: the Spatial Matrix changed suddenly (Δ). The connection to the outside was severed (Ω — window vacuum closed). ANS produces a pressure response.

Gemini wrote this sentence — and it's the best sentence in the scene.

And it doesn't know why it wrote it.

References

• Bulut, L. (2026). Architectural Framework of Objective Projection. Zenodo. https://doi.org/10.5281/zenodo.18689179
• Bulut, L. (2026). The Ng Operator: Mathematical Formalization of Narrative Gravity. Zenodo. https://doi.org/10.5281/zenodo.19275490
• Bulut, L. (2026). Beyond Eliot: From Objective Correlative to Objective Projection. Zenodo. https://doi.org/10.5281/zenodo.19390047 | SSRN: 6510742
• Bulut, L. (2026). Sn Measurement Protocol v1.0. Zenodo. https://doi.org/10.5281/zenodo.19410663
• Eliot, T.S. (1919). Hamlet and His Problems. The Sacred Wood. Methuen.