There are events the public record cannot close.
The Kennedy assassination is the canonical case. We are interested in why.
Most events decay. The record settles, attention disperses, interpretation stabilizes. A consensus version emerges and persists, not because it is correct, but because the alternatives cost more to maintain than they return.
Closure is not the same as truth. The framework cares about stability, not correctness.
Closure is not a truth condition. It is a stability condition. A false account can close. A true account can fail to close. The framework is concerned with stability, not with veracity.
A small number of events do the opposite. They accumulate. New interpretations arrive faster than old ones decay. The record grows denser over time, not thinner. The event does not close. It compounds.
The framework borrows vocabulary from physics and information theory. It does not borrow mathematics. The events behave as if they have mass. The behaviors are observable. The mechanism is speculative. Its value is not in being true. It is in being useful.
Some events accumulate variants faster than the system can absorb them.
Sixty-three years after the event, the Kennedy record is larger than it was at one year, at five years, at thirty. The same pattern appears around the September 11 attacks, the Epstein case, the death of Princess Diana, the question of COVID-19's origins, the Roswell incident, and a small handful of others. Each accumulates. None closes.
The pattern is specific and observable. It is not the same as being remembered. It is not the same as ongoing relevance. It is not the same as scandal or controversy.
What does not accumulate
| Event type | Why it does not accumulate |
|---|---|
| Major sporting event | High volume but resolves cleanly. The score is known. |
| Natural disaster | High attention but low interpretive diversity. |
| Ordinary political scandal | Decays after institutional or legal closure. |
| Viral hoax | High spread but shallow persistence. |
| War | Too broad to count as a single event. Specific events within it can accumulate. |
Importance is not enough. Media coverage is not enough. Ambiguity at the moment is not enough. Something else is operating.
Old patterns determine which modern events can become massive.
Certain figures and events recur across cultures because they fit structural slots. The slain king. The hidden hand. The captive princess. The visitor from elsewhere. The fallen citadel. The vanishing. These are the shapes the human imagination uses to organize meaning around traumatic public events.
Kennedy is the slain king at noon. The September 11 attacks are the falling citadel. The Epstein case is the hidden conspiracy of power. Princess Diana is the captive princess. Roswell is the visitor. Amelia Earhart is the vanishing.
Myth used to close stories. Now it just opens them.
These slots historically performed a stabilizing function. The mythological function was to give a culture a way to metabolize the event and reach closure. We are arguing that the function has broken. The pull remains. The closure does not.
Three terms, doing most of the work.
Four types of unresolvability
Kennedy exhibits all four. Epstein exhibits all four. COVID origins exhibits evidentiary, institutional, and strategic. The mix matters because different types respond to different interventions.
Past a certain point, the event cannot settle.
At some point the accumulating mass crosses a threshold past which the system can no longer compress the event into a stable record. We call that threshold interpretive criticality.
The closure function operates only in cultures with trusted authorities, slow variant distribution, and shared frameworks for what a settled account looks like. Modern conditions provide none of these.
The Kennedy assassination crossed long ago. The Epstein case crossed recently. The September 11 attacks are in a slow transit. The pandemic origins question is approaching.
The field reaches for what would resolve the contradiction.
At criticality, the field generates an attempt to produce the missing source. Whatever is most absent from the record gets pushed toward presence. We name this temporal extrusion.
The hard reading: the field literally produces the missing subject. That reading is the central conceit of an accompanying work of fiction.
The softer reading: extrusion is already happening functionally. AI systems generate the missing testimony. Synthetic media produce the missing footage. Deepfakes produce the missing interview. The contest about authenticity becomes part of the mass.
Even the missing source, when produced, becomes another voice in the contest.
The field that has crossed into criticality is in equilibrium. The equilibrium is not consensus. It is contested interpretation. The field has a structural stake in its own continuation.
Information environments do not want resolution. They want equilibrium, and the equilibrium they have reached for these events is the contest itself.
We name this the narrative recoil. It prevents any single operation from collapsing the accumulated load. The recoil is not driven by any agent. It emerges from the structure.
Correction is contribution. The field cannot be drained, only fed.
Four operations the field performs at criticality.
Three are emergent behaviors of the field. The fourth is what happens when agents try to act on the field.
The four are not mutually exclusive. The framework allows the analyst to ask, for any given event, which operations are most active.
Closure work is what the analyst can actually do.
The closure window opens when the event occurs and closes when the system crosses interpretive criticality. After the window closes, the recoil takes over and conventional intervention becomes counterproductive.
Closure work is preemptive structuring. Establish trusted authorities before competing ones can be improvised. Frame a consensus baseline before variants proliferate. Acknowledge the archetypal slot directly. Provide a complete official account quickly.
If the framework is approximately correct, several postures need adjustment.
01 · Correction does not work on high-mass events
Fact-checking, file releases, official commissions do not discharge mass. They add to it. This is not a failure of execution. It is a structural property of high-mass fields.
02 · Disinformation succeeds structurally even when its claims fail
Bad actors do not need their lies believed. They need the contest to continue. The campaign does not have to win. It has to deny closure.
03 · New tools are throughput accelerators, not new threats
Generative AI increases the rate at which variants enter the field. It does not change the mechanism. The intervention point is not the tool. It is the conditions under which the field reaches criticality.
04 · The typology gives analysts a diagnostic question
For any contested event: which of the four operations is most active? The diagnostic has two axes. Which operation is active. How much time remains.
05 · Some events have crossed past intervention
Redirect effort from correction to characterization. Map the field. Track the variants. The goal is to understand what its continued operation does to adjacent information environments.
Six questions for any contested event.
The diagnostic does not require quantitative answers. Each question can be answered categorically. The pattern of answers indicates how far the event has progressed.
The diagnostic does not measure. It locates.
The framework is speculative.
It borrows vocabulary from physics and information theory to describe phenomena those fields do not address. We claim that narrative fields behave as if they have mass, in ways that are observable, patternable, and useful to track.
Several limits should be acknowledged. No formal measurement of narrative mass. The four operations are not exhaustive. The closure window has no direct clock. The threshold of interpretive criticality is identified retrospectively more easily than prospectively.
The novella inhabits a speculative premise fully. The paper does not. The two share vocabulary and not much else.
The framework's value is not in being true. It is in being useful.
A high-mass event cannot always be corrected. But it can be mapped. That is where serious work begins.