THE CONSPIRACY QUESTION IS NOT THE OSWALD QUESTION.

There is no single “JFK conspiracy question.” There is a stack of questions that history, popular culture, government secrecy, institutional embarrassment, and sixty years of argument have welded into one noisy object.

That makes bad reasoning almost automatic. Evidence that supports one proposition gets borrowed by another. Evidence that weakens one claim gets treated as if it annihilated the entire opposing worldview. A newly released CIA document can be waved around as proof of a second shooter even if it says nothing about Dealey Plaza. A technically feasible rifle shot can be presented as proof that no one else knew anything.

The repair is straightforward: separate the propositions.

Question 1: Did Lee Harvey Oswald fire at Kennedy?

This is the shooter-authorship question. It concerns the rifle, cartridge cases, firing location, ballistics, witness and photographic evidence, Oswald's movements, possession of the weapon, and related forensic material.

The Warren Commission concluded that Oswald was the assassin. The later House Select Committee on Assassinations also concluded that Oswald fired three shots at Kennedy and that two hit him, including the fatal shot.

That agreement is important because the two investigations did not agree on everything else.

Question 2: Was there another shooter?

This is a separate proposition. The HSCA answered it differently from the Warren Commission because acoustic analysis persuaded the committee that there was a high probability of another gunman firing from the grassy-knoll area.

That conclusion later suffered a major technical blow. In 1982, the National Research Council's Committee on Ballistic Acoustics unanimously concluded that the acoustic analyses did not demonstrate a grassy-knoll shot and that there was no acoustic basis for the claimed 95 percent probability.

That weakens the specific acoustic route to a second gunman. It does not logically prove that a second gunman could not have existed by some other evidentiary path. It means the famous acoustic support cannot simply be cited as settled science.

Question 3: Did anyone assist Oswald?

A person can be the only shooter and still participate in a conspiracy. Assistance could involve planning, logistics, intelligence, transportation, money, weapons, surveillance, communications, or post-event escape arrangements.

Evidence relevant here is often completely different from evidence about bullet trajectories. Communications records matter. Associations matter. Financial records matter. Travel and contact histories matter. The absence of a second rifle does not answer whether another person encouraged, facilitated, or coordinated the act.

Question 4: Did anyone know beforehand?

Foreknowledge is narrower than participation and broader than innocence. An intelligence service could have collected warning information and failed to act. An informant could have heard violent intent. An associate might have known something without helping. A bureaucracy could possess fragments that only become meaningful after the event.

This is where declassified records become especially important. A document showing that an agency monitored Oswald more closely than previously admitted may change our understanding of institutional knowledge even if it contains no evidence that the agency ordered the assassination.

Question 5: Did institutions conceal, distort, or withhold information afterward?

This is perhaps the most commonly conflated category.

Government agencies can hide mistakes, intelligence methods, embarrassing surveillance, illegal programs, foreign contacts, sources, or bureaucratic failures without having committed the underlying murder. Conversely, proving an institutional cover-up of some fact does not tell us automatically what that fact was covering.

“They lied about X” is evidence that they lied about X. It is not a magic bridge to “therefore they killed Kennedy.” The bridge requires its own evidence.

The government's own record demonstrates the distinction.

The HSCA is almost a textbook example. Its summary findings said Oswald fired the shots that struck Kennedy. It also said Kennedy was probably assassinated as a result of a conspiracy, while acknowledging it could not identify the other gunman or the extent of that conspiracy.

That position only looks contradictory if “Oswald shot Kennedy” and “there was a conspiracy” are incorrectly treated as mutually exclusive statements.

The new records should be scored proposition by proposition.

The National Archives released more than 80,000 pages of previously classified JFK records in March 2025 and has continued publishing material, including another 11,022 pages on January 30, 2026. The total JFK collection spans millions of pages and many media types.

That volume creates an obvious failure mode: every interesting document becomes “new JFK bombshell.” A better system would maintain a living evidence ledger.

For each new record, ask:

Does it change the shooter evidence?
Does it support or weaken a second-shooter theory?
Does it show assistance?
Does it change what an agency knew beforehand?
Does it document later concealment or institutional misconduct?
Does it change nothing about the assassination itself but improve historical context?

Most documents will land in the last few categories. That is not disappointing. History is usually made of incremental changes to the map, not daily explosions.

A conspiracy theory should be modular enough to fail.

This is where serious investigation parts company with belief maintenance. If acoustic evidence fails, remove or downgrade the acoustic branch. If a witness is contradicted, reduce that witness's weight. If an agency document shows embarrassing secrecy but no operational connection to murder, record the secrecy without laundering it into homicide evidence.

A theory that cannot lose pieces when evidence turns against them is not an investigation. It is a religion with footnotes.

Our next step

Cyberdelia's JFK desk should build that ledger publicly. Every major claim gets a status: strengthened, weakened, unchanged, unresolved, or malformed. New records get attached to the claims they actually bear on. Readers should be able to see the inference rather than being asked to inherit it.

That is slower than shouting “case closed.” It is also how one eventually learns what the case contains.