> ## Content Index > Fetch the complete content index at: https://otherwise.ghost.io/llms.txt > Use this file to discover other available public pages before exploring further. # When Catastrophe Becomes Conspiracy - URL: https://otherwise.ghost.io/when-catastrophe-becomes-conspiracy/ - Published: 2026-09-10T16:43:41.000Z - Updated: 2026-09-10T16:43:41.000Z - Description: September 11, October 7, and the Strange Human Need to Turn Failure into Intention - Author: Khaled Salih - Tags: Newsletter, #Migrated-1792244264359, #Import 2026-10-17 14:38 [10 September 2026](https://x.com/khaledsalih2026/status/2097999675034935426?s=20&ref=otherwise.ghost.io) # Note What if conspiracy theories do something stranger than simply get catastrophes wrong? September 11 and October 7 produced the same haunting question: **How could institutions built to prevent such attacks fail so catastrophically?** Real warnings existed. Real assumptions failed. Real institutions made grave errors. But then an extraordinary transformation can occur: ***They failed to understand → They knew → They allowed it → They wanted it.*** In this essay, I develop a new conjecture: conspiracy thinking can perform an **epistemological rescue**. It turns distributed institutional failure into centralized hidden intention. The world becomes darker—but strangely easier to understand. Someone, at least, was in control. The paradox is that what presents itself as radical skepticism can become an **anti-correction architecture**, diverting criticism away from the assumptions, decisions and institutions that actually failed. My formulation is: **Conspiracy theory can become a protected explanation of protected error.** The alternative is not trust in authority. It is something much harder: **Find where the error lived. Discover what protected it. Preserve criticism. Keep Backspace available.** Because when institutions cannot press Backspace, reality eventually does it for them. **When Catastrophe Becomes Conspiracy** September 11, October 7, and the Strange Human Need to Turn Failure into Intention --- # When Catastrophe Becomes Conspiracy September 11, October 7, and the Strange Human Need to Turn Failure into Intention There is something about catastrophe that makes ordinary explanations feel insufficient. On September 11, 2001, nineteen hijackers penetrated the security architecture of the most powerful country in the world. They turned passenger aircraft into weapons, destroyed the World Trade Center towers, struck the Pentagon, and killed almost three thousand people. On October 7, 2023, Hamas penetrated the security architecture of one of the most heavily surveilled borders in the world. Its attackers overran military positions, entered Israeli communities, murdered civilians and soldiers, and abducted hundreds of people into Gaza. The events are different. Their histories are different. Their perpetrators, political contexts, military circumstances and consequences are different. But they produced a remarkably similar question: ***How could this possibly have happened?*** That is a good question. Indeed, it is the question from which any serious investigation should begin. The problem begins when another question silently replaces it: ***Could they really not have known?*** And from there: **Perhaps they did know.** Then: **Perhaps they deliberately did nothing.** And finally: **Perhaps they wanted it to happen.** Something extraordinary has occurred between the first question and the last. Failure has become intention. That transformation interests me. I want to make a conjecture. # Start With the Failure I do not begin from the assumption that conspiracy theories are produced simply by stupid, irrational or gullible people. That explanation is too easy. Intelligent people believe conspiracy theories. Educated people believe them. People capable of sophisticated criticism believe them. And governments really do conspire. Intelligence agencies really do conduct covert operations. Officials really do lie. Institutions really do conceal embarrassing failures. Political leaders really do exploit crises they did not create. So the interesting question cannot be: ***Why don’t people simply trust the authorities?*** They should not simply trust the authorities. Neither should they simply distrust them. The question I want to ask is different: ***What happens to our explanations when institutions really have failed?*** Here September 11 and October 7 become particularly revealing. There were real warnings before September 11\. The American intelligence system possessed fragments of highly consequential information. Agencies failed to connect information effectively. Threat assumptions proved inadequate. The 9/11 Commission eventually described failures of imagination, policy, capabilities and management. October 7 presents an equally disturbing problem. Israel possessed the Hamas attack document commonly known as the “Jericho Wall” plan well before the assault. Warning signs existed. Yet the prevailing assessment treated the scale of attack described in the plan as beyond what Hamas was likely or able to execute. Subsequent Israeli investigations have documented profound failures in intelligence, preparedness and military response. These facts matter. They must not be explained away merely because conspiracy theorists use them. But neither do they establish conspiracy. That distinction is where I want to begin. # The Missing Arrows Consider four propositions. **There was warning.** **The warning was mishandled.** **Someone negligently failed to act on the warning.** **Someone deliberately allowed the attack to happen.** They are not four formulations of the same proposition. They are four different conjectures. And the evidential distance between them is enormous. Yet conspiracy reasoning frequently transforms them into a continuous sentence: They had warnings, therefore they knew; they knew, therefore they deliberately failed to act; they deliberately failed to act, therefore they wanted the attack. Look carefully. The important part of that sentence is not the claims. It is the word: **therefore.** Each *therefore* conceals an explanatory bridge. And each bridge requires evidence. This suggests a rule I want to add to the Open Correction Framework: **Do not audit only the propositions in an argument. Audit the arrows between them.** I call this the **Inference-Bridge Test**. A conspiracy theory may contain many true facts and still be a bad explanation because the causal bridges connecting those facts have never been established. This is why anomaly accumulation is not enough. Ten unanswered questions do not automatically constitute one answer. One hundred anomalies do not manufacture a mechanism. # Perhaps the Most Frightening Explanation Is Failure Why are we tempted to make the jump? Here is my conjecture. Perhaps conspiracy sometimes gives us something psychologically reassuring that institutional failure does not. Suppose the United States government secretly organized September 11. That would be morally monstrous. Suppose Israeli leaders knowingly permitted October 7. That too would be morally monstrous. But notice what both explanations restore: **competence.** Someone knew. Someone understood. Someone anticipated what would happen. Someone remained in control. The world becomes evil—but intelligible. Now consider the alternative. Thousands of intelligent people can work inside extraordinarily sophisticated institutions possessing enormous technological, military and informational capacities—and still misunderstand reality catastrophically. No one needs to want the catastrophe. No one needs to coordinate the failure. The system can simply be wrong. That possibility is morally less sinister. But epistemically it may be much more frightening. I call this the **Competence Restoration Conjecture**: **Conspiracy can transform an intolerable story of distributed fallibility into a more psychologically manageable story of centralized competence. The institution did not really fail. Someone secretly succeeded.** The conspirator may therefore be evil. But at least someone was driving the car. # “They Knew” There is another problem hidden inside that phrase. **They knew.** Who is *they*? An intelligence officer? A field analyst? The FBI? The CIA? The White House? Shin Bet? Military Intelligence? A surveillance unit? The IDF chief of staff? The prime minister? A bureaucracy does not possess a single mind. One person possesses information A. Another possesses B. A database contains C. A field officer reports D. A senior official receives a compressed summary containing parts of A and C. Another analyst interprets D through an existing model and downgrades it. The information may exist inside the institution without becoming institutional knowledge capable of producing action. So asking: ***Did America know?*** or: ***Did Israel know?*** can already contain an epistemological mistake. The proper questions are harder. **Who knew what?** **When?** **With what confidence?** **Who received it?** **How was it classified?** **Which explanation was used to interpret it?** **Who had authority to act?** **What happened to contradictory assessments?** **Which threshold had to be crossed before information became action?** This gives me another concept: ***The Knowledge-Aggregation Error*** **Conspiracy reasoning often converts information distributed across an institution into knowledge supposedly possessed by a single collective mind.** Once that fictional mind has been created, intention becomes much easier to infer. “They” knew. “They” chose not to act. “They” therefore wanted the result. But the institution may never have possessed the integrated knowledge attributed retrospectively to it. # The More Interesting Parallel This is where September 11 and October 7 become much more interesting to me. The important parallel may not be conspiracy at all. It may be **explanatory protection**. Before September 11, American institutions possessed information about al-Qaeda and the threat it posed. But institutions were still operating with inherited assumptions about aviation security, hijacking, intelligence boundaries and what terrorists were likely to attempt. The relevant information existed. The explanatory architecture required to interpret it correctly did not. Something similar appears to have happened before October 7. Israel knew a great deal about Hamas. That may have been part of the problem. The prevailing conception increasingly treated Hamas as deterred, governable through incentives, interested in preserving its rule in Gaza, and unlikely to undertake an operation of the magnitude eventually carried out. Information contradicting that explanation arrived. But information never arrives naked. It arrives inside explanations. And the governing explanation tells us what the information means. That leads me to what I regard as the strongest conjecture emerging from comparing these two catastrophes. **The Protected Model Catastrophe** **Catastrophic surprise can occur not because an institution possesses too little information, but because it possesses a protected explanation through which available information is repeatedly interpreted. The institution may therefore possess many of the relevant facts while remaining epistemically incapable of understanding what those facts imply.** This is different from intelligence failure in the ordinary sense. It is an **explanatory failure**. The facts are present. But they are imprisoned inside the wrong explanation. That is much more dangerous. Because collecting more information may not solve it. # Intelligence Is Not Rationality I have repeatedly argued that **intelligence is not rationality**. September 11 and October 7 allow that distinction to be moved from the individual to the institutional level. The United States did not lack intelligence institutions. Israel certainly did not lack them. Both possessed formidable technical capacities, experienced analysts, surveillance systems, military power and accumulated knowledge about their adversaries. Yet none of those things guarantees rationality. Indeed, an uncomfortable possibility emerges: **An institution can become better at collecting intelligence while becoming worse at criticizing the explanation through which that intelligence is interpreted.** More data may not correct the model. It may strengthen confidence in it. More sensors. More intercepts. More analysts. More algorithms. More surveillance. More expertise. And still: the wrong explanation. This produces what I call the **Epistemic Supremacy Trap**. The institution gradually acquires confidence not merely that it possesses information, but that it **understands the adversary**. We know Hamas. We know al-Qaeda. We know what they can do. We know what they cannot do. We know what they want. We know what will deter them. At that moment intelligence can become an obstacle to rationality. Because the anomaly is no longer allowed to criticize the explanation. The explanation explains away the anomaly. Until reality does the criticizing. At catastrophic cost. # Then Something Strange Happens The conspiracy theorist arrives after the catastrophe and commits a remarkably similar error. The institution had said: **We understand the enemy.** Reality proved otherwise. The conspiracist now says: **I understand the institution.** The first protected model was the institution’s model of its adversary. The second protected model becomes the conspiracist’s model of the institution. Different explanations. Same epistemological temptation. Certainty. # October 7 Requires Particular Discipline October 7 is especially important because some of the material used by conspiracy theories refers to genuine problems. There were warnings. There was the Jericho Wall plan. There were serious failures of interpretation. There were devastating failures of military preparedness and response. Israeli fire killed some Israelis during the fighting. Questions about orders intended to prevent abductions deserve investigation where evidence supports them. None of this should be hidden. To hide genuine failure because conspiracists might exploit it would itself create protected error. But the reverse move is equally irrational. The fact that A is true does not make B true. We therefore need an **Inference Ladder**: 1. **Warning existed.** 2. **Warning was incorrectly interpreted.** 3. **Institutional assumptions contributed to the error.** 4. **Officials may bear professional, command or political responsibility.** 5. **Israeli actions during the response caused Israeli deaths in some cases.** Then come much stronger propositions: 1. **Israeli forces operated under a general policy deliberately sacrificing Israeli civilians.** 2. **Israeli leaders deliberately permitted Hamas to carry out the attack.** 3. **Israel secretly participated in producing October 7.** Moving from 1 to 8 is not investigation. Each level creates a new evidential burden. The lower propositions do not automatically validate the higher ones. So another Open Correction Framework rule follows: **The existence of a lower-level scandal never cancels the evidential burden for a higher-level conspiracy claim.** # Benefit Is Not Intention Another arrow needs criticism. September 11 transformed American politics. It generated wars, security institutions, surveillance powers, budgets and political opportunities. October 7 transformed Israeli politics, Gaza, regional strategy and the international environment. Political actors can exploit catastrophes. Governments can use them. Institutions can expand their powers because of them. None of that establishes that they created the event from which they later benefited. The reasoning: **X benefited from Y, therefore X caused Y** is extraordinarily weak. History is filled with people exploiting events they neither predicted nor produced. So I would add another rule: **Never infer pre-event purpose from post-event utility without an independent causal mechanism connecting the two.** We must distinguish: **made it happen** from **let it happen** from **failed to prevent it** from **capitalized on what happened.** These are different explanations. Collapsing them does not increase criticism. It destroys precision. # Where Does the Error Go? Now I reach the question at the center of the Open Correction Framework: ***Where does the error go?*** After September 11, it could go into: intelligence sharing; watchlisting; aviation security; threat assumptions; CIA-FBI coordination; institutional imagination; executive decision-making; military preparedness; and later political decisions made in response to the attack. After October 7, it could go into: the conception of Hamas; intelligence assessment; border architecture; force deployment; political assumptions; command structures; warning escalation; military response; hostage doctrine; civil-military relations; and political responsibility. Those are addresses. Error can be sent there. Institutions can be changed. People can be removed. Assumptions can be revised. Procedures can be redesigned. Now compare: **They wanted it to happen.** The accusation sounds more radical. But where does the error go? Into an invisible total actor. The specific institutional failures begin disappearing. The conspiracy theory may therefore sound more critical while becoming **less capable of correction**. That is the paradox that interests me most. # Conspiracy as Anti-Correction I can now state my central conjecture more precisely. ***The Correction-Displacement Conjecture*** **Conspiracy theories flourish especially powerfully where genuine institutional failure has produced unresolved anomalies, broken trust and insufficient correction. They convert distributed fallibility into centralized hidden intention because intentional evil is easier to narrate than systemic epistemic failure. But in doing so they commit a second-order error: criticism is displaced from the assumptions, decisions, explanations and institutions that actually failed toward an increasingly total and unfalsifiable hidden actor.** This means conspiracy theory can become something more interesting than misinformation. It can become an **anti-correction architecture**. It begins with legitimate criticism. Why was this warning ignored? Why did this institution fail? Why did the response take so long? Why did officials say something afterwards that proved false? Those questions should remain open. But gradually the questions become an answer. And the answer becomes protected. No evidence can defeat it. Missing evidence proves concealment. Contradictory evidence proves fabrication. Official documents prove coordination. Missing documents prove destruction. A witness confirms the theory. A witness denying it has been compromised. The theory has acquired a remarkable property: **it cannot lose.** That is not explanatory strength. It is closure. # A Protected Explanation of Protected Error This produces a structure I had not previously formulated this way. The first failure is institutional: **A bad explanatory model becomes protected.** Reality eventually destroys it. The catastrophe occurs. Society then tries to explain the catastrophe. A second protected explanation can arise: **The institution did not fail. It secretly intended the outcome.** We therefore obtain: **Protected institutional explanation** **→ catastrophic falsification** **→ conspiracy explanation** **→ protection of conspiracy explanation** **→ blocked correction.** The conspiracy theory becomes: **a protected explanation of protected error.** That, I think, is the deeper architecture connecting September 11 and October 7. And it has reach far beyond either case. # The Missing-Anomaly Test There is another test I want to introduce. Conspiracy theories are fascinated by anomalies that appear to suggest coordination. But what about anomalies pointing in the opposite direction? What would we expect if an institution were genuinely surprised? Confusion. Contradictory information. Communication breakdown. Improvised decisions. Friendly fire. Misidentification. Slow response. Competing commands. People acting without complete knowledge. Institutional chaos. Both September 11 and October 7 contain enormous amounts of precisely this material. So ask: **Does the conspiracy explanation explain the incompetence, or does it simply redescribe evidence of incompetence as evidence of extraordinary competence?** This is the **Missing-Anomaly Test**. A theory that counts every apparent sign of coordination but refuses to count evidence of chaos is selecting its evidence directionally. **The Directionality Test** Everyone makes mistakes. Investigators make mistakes. Journalists make mistakes. Governments make mistakes. Conspiracy theorists make mistakes. That alone tells us little. But suppose every ambiguity moves in one direction. Every unknown supports the conspiracy. Every mistake enlarges it. Every correction merely changes its form. Every absence of evidence proves concealment. Every contradiction proves deception. Then the important fact is not that errors occurred. It is their **directionality**. Ask: ***Can any new fact weaken the conjecture?*** If the answer is no, we no longer possess a conjecture. We possess an immune system. # The False Binary There is also a choice-space problem. The public is frequently offered two doors: **Official story.** or **Conspiracy.** I reject both-door architecture. The fact that an official account contains errors does not validate an inside-job theory. The fact that a conspiracy theory contains errors does not validate every official account. There are many possibilities between those doors: mistaken intelligence; bureaucratic fragmentation; negligence; groupthink; protected assumptions; miscommunication; institutional rivalry; political interference; localized misconduct; friendly fire; later concealment of embarrassment; post-event exploitation; and combinations of them. This is exactly why the Choice-Space Audit matters. Before deciding which explanation wins, ask: ***Who created the menu?*** “Official story versus conspiracy” may itself be the first bad explanation. # Conspiracy and Backspace A good conjecture must be capable of dying. I should be able to say: **Here is my explanation.** **Here is why I think it is better.** **Here is the mechanism.** **Here is what remains uncertain.** And crucially: **Here is what would make me abandon it.** That last sentence is Backspace. A protected conspiracy theory reverses the architecture. Evidence can enter. Evidence cannot leave. Everything becomes confirmation. It is an epistemic one-way door. That is why I do not reject conspiracy claims because conspiracies are impossible. Of course conspiracies occur. I reject particular conspiracy explanations when they have not earned the causal bridges they require. Capacity is not occurrence. Suspicion is not mechanism. Warning is not intention. Negligence is not orchestration. Benefit is not causation. Secrecy is not proof. And unanswered questions are not answers. # What Would Change My Mind? The Open Correction Framework must impose this requirement on me too. What would cause me to reconsider the present explanation of September 11 or October 7? Direct contemporaneous orders deliberately permitting the attack. Authenticated communications establishing prior coordination with the perpetrators. Documented suppression of precise actionable warnings specifically for the purpose of ensuring that the attack succeeded. Operational or financial coordination tied directly to the attack. Independent documentary and testimonial evidence establishing a deliberate facilitation mechanism. Such evidence would matter enormously. I would have to press Backspace. That is precisely the point. The present conclusion must never become: **A conspiracy is impossible.** It should be: **The evidence currently available does not justify moving from catastrophic institutional failure to deliberate orchestration.** That is a Lock. Not a sacred conclusion. # The Real Scandal There is a final irony. People sometimes think that rejecting conspiracy means defending institutions. I reach almost the opposite conclusion. The non-conspiratorial explanation may be more devastating. September 11 tells us that one of the most powerful states in history could possess formidable intelligence capabilities and still fail to understand what was coming. October 7 tells us that one of the world’s most sophisticated security states could watch an adversary intensively, possess highly consequential information about its plans, and still misunderstand what that adversary was prepared to do. That is not reassuring. It is terrifying. Because conspiracy gives us a villain. Fallibility gives us ourselves. # A Better Conjecture? I would therefore formulate the argument this way: **Conspiracy theories become especially attractive after catastrophic institutional failure because they perform an epistemological rescue. They transform distributed failure into centralized intention, fragmented information into unified knowledge, incompetence into secret competence, post-event benefit into pre-event purpose, and unresolved anomalies into a total explanation. In doing so, they can divert criticism from the very explanations and institutions that reality has shown to be defective.** The deeper sequence is therefore: ***Protected model → ignored or misread criticism → catastrophe → shattered trust → conspiracy explanation → protected conspiracy → displaced correction.*** This is why conspiracy theory is not merely a problem of false belief. It is a problem of **where error goes**. **Before Catastrophe Presses Backspace** September 11 and October 7 should therefore teach us something more important than either “trust your institutions” or “never trust your institutions.” Both are bad epistemology. Institutions are fallible. Critics are fallible. Intelligence agencies are fallible. Governments are fallible. Journalists are fallible. Skeptics are fallible. I am fallible. The rational question is never who deserves immunity from criticism. It is: ***Which explanation survives criticism—for now?*** And then: ***What architecture allows us to discover that it is wrong before the price of correction becomes catastrophic?*** That is the question September 11 should have forced upon the United States. It is the question October 7 should force upon Israel. And it is the question conspiracy theory too often prevents us from asking precisely enough. The deepest lesson is therefore not that somebody secretly controls history. It is almost the opposite. **Nobody controls knowledge that has not yet been created.** Highly intelligent institutions can misunderstand reality. Power cannot abolish fallibility. Surveillance cannot replace explanation. Information cannot correct a model that has acquired protection from criticism. And intelligence cannot guarantee rationality. The task is therefore not to create institutions that will never be wrong. There are no such institutions. The task is to create institutions in which error can travel upward—into the assumption, the explanation, the doctrine, the decision, the leader and the architecture—before it travels downward into thousands of dead human beings. That is the alternative to conspiracy. Not trust. **Correction.** Not certainty. **Criticism.** Not an omniscient hidden hand. **Fallible human beings living inside institutions that must remain capable of pressing Backspace.** Because when institutions cannot press Backspace, reality eventually does it for them. And reality does not use an eraser.