The Theories

Listen to this chapter, read by the author →

One at a time

‡The theory goes like this. A shooter waited inside the storm drain in the curb on Elm Street, fired through the opening at road level, and slipped away underground through the drainage system. It is neat and it explains why nobody saw a gunman. Then you go and look at it. The sight line from inside is low and blocked by the curb itself, the drain does not connect to anything a person could crawl out through, and nothing has ever been found in it.^{1, 2, 3}

‡In the weeks before Dallas, people reported seeing Lee Harvey Oswald in places he was not. A rifle range, a car showroom, driving a car he could not drive. The theory is that somebody was impersonating him to build a trail. It is one of the more serious claims, because the sightings are real reports by real people. It is also what you get when a face becomes the most published photograph in the world and everyone searches their memory of the past month.^{4, 5}

‡In the films there is a woman in a headscarf standing on the grass, holding what looks like a camera, pointed straight at the car. She never came forward and no film from her position has ever surfaced. One woman later claimed to be her and said her film was taken by federal agents. Her account has serious problems, including details she could only have known from published sources. So the woman is real. Everything after that is not established.^{6, 7}

‡Enlarge Mary Moorman’s Polaroid far enough and a shape appears behind the stockade fence that some people read as a man in a police uniform, firing. It was named Badge Man in 1982 and it has never gone away. The problem is the print itself: at that magnification you are looking at the grain of a 1963 instant photograph. Push any image far enough and it will show you a face. That is a property of grain, not of fences.^{8, 9}

‡One theory holds that the man with the open umbrella on a clear day fired a paralysing dart, freezing the President in his seat for the shooters. It was even demonstrated to Congress with a working device. What Congress also got, in 1978, was the man himself. He turned up, brought the umbrella, and explained it was a heckle about the President’s father. The dart gun existed. The evidence that this umbrella was one does not.¹⁰

‡This one spreads faster than any other because it looks like proof. In a poor copy of the Zapruder film, a flash of light appears near the driver’s shoulder at the moment of the head shot, and it is read as a pistol. In a clean transfer it resolves into sunlight on the hair of the agent in the passenger seat. Bill Greer had both hands on the wheel. This is not an interpretation problem. It is a video quality problem.^{11, 12}

‡Over sixty years this killing has been attributed to the Mafia, the CIA, the KGB, Cuba, anti-Castro exiles, the FBI, the military, Texas oil, and the man who became President. Each theory has serious people behind it and a genuine motive to point at. What none of them has is a document, a confession that checks out, or a piece of physical evidence from that plaza. They cannot all be right. They can all be unproven, and they are.^{13, 14}

‡One branch of theory holds that the evidence was changed between Dallas and Washington to disguise where the shots came from. It rests on real discrepancies. Doctors in Texas and doctors in Maryland described things differently, and those differences are documented. The leap is from disagreement to alteration. Two sets of doctors, hours apart, under enormous pressure, describing the same injuries differently is the most ordinary thing in medicine.^{15, 16, 17}

‡Two days later the only suspect was shot dead in a police basement by a nightclub owner, live on television. If you wanted one fact to make a whole country suspicious forever, that is the fact. The theory is that Ruby was sent. What is against it is the sheer improbability of the plan: it depended on a police transfer running late, on Ruby happening to be nearby, and on him walking down a ramp nobody was watching.^{18, 19}

‡Here is the honest reason none of this ends. Agencies did lie to the investigators about other things. Files were withheld for decades for reasons that had nothing to do with the plaza. So the people who say the government has not been straight about this are right, just not about the six seconds. Real institutional dishonesty about the surrounding history is the fuel. It keeps every theory alive, including the ones with nothing in them.^{20, 21, 22}

The mob case

In 1979 a Congressional committee wrote that organized crime as a national body was not involved in the assassination, and then wrote the other half of the sentence: that the evidence did not preclude individual members having been involved. That is not a theory somebody had. It is a finding of the United States House of Representatives, and it names two men.²³

Carlos Marcello ran New Orleans. In 1961 the Attorney General’s department had him seized without warning and flown to Guatemala, no hearing, no chance to pack. He was back within months, and he had a specific and personal reason to hate the family that did it. Motive in this case is usually assumed. In his it is documented, and the government created it.^{14, 24}

Santo Trafficante had run casinos in Havana until Castro took them and briefly jailed him. He is one of the criminals the CIA then recruited to try to kill Castro, alongside Johnny Roselli. So the same man appears in two places in this story: in the mob theory, and in a government operation the Warren Commission was never told existed. That overlap is the reason the mob question could not be cleanly answered.^{14, 25}

†A businessman named Edward Becker said he was at a farmhouse in Louisiana in 1962 when Marcello used a Sicilian expression about taking a stone out of his shoe, and said that going after the brother was useless. You cut off the dog’s head, not its tail. Marcello denied it. The Bureau questioned Becker’s reliability. The remark has been quoted for fifty years and has never been corroborated by anybody who was in the room.^{14, 26}

The strongest thread is not a quotation. It is a telephone record. In the months before November the Dallas nightclub owner who would shoot the prisoner made a rising number of calls to men connected to both of those organizations. The Warren Commission had described him as a man with no significant underworld ties. Fifteen years later the House committee went through the toll records and found that description was simply wrong.^{27, 28}

†This is the best-supported conspiracy case there is. It has motive that the government itself manufactured, two named men, a Congressional committee declining to rule them out, and a shooter with the right phone calls. What it has never had is one document, one participant who confessed and checked out, or one piece of physical evidence from that plaza. It is the most serious theory in the field and it is still, after sixty years, unproven.^{14, 29}

The myths

†You will hear that the Carcano was junk and the shot was impossible. It is a cheap surplus rifle, but it is a military weapon that worked, and marksmen who have tested it get consistent results at that range. The distance to the car at the final shot is about 265 feet, downhill, at a target moving in a straight line at eleven miles an hour. That is not an easy shot. It is also not an extraordinary one.^{30, 31, 32}

†The claim that the suspect could not shoot comes up constantly. His Marine Corps records show him qualifying as a sharpshooter at one point and scraping the lowest rating at another. So the honest summary is that he was a trained rifleman who was inconsistent, tested years earlier, on different weapons, at different distances. People quote whichever of those two numbers suits them.^{33, 34}

†The single most quoted argument for a shot from the front is that the film shows the head go back. It does. What is contested is what that means: a bullet is not a shove, and physicists have offered several mechanisms for rearward motion after a rear entry, none of which command universal agreement. This is one where the honest answer is that reasonable people who understand the physics still disagree.^{35, 16, 36}

The phrase is about a path, not an object. The claim is that one bullet could not have made seven wounds in two men. The answer is that the two men were not sitting the way people picture them. The Governor was on a jump seat, lower and set inboard, not directly in front. Line them up as they actually sat and the path is close to straight. Whether it happened that way is arguable. Whether the geometry forbids it is not.^{37, 38}

†The storm drain theory is durable because it sounds clever. Walk it, though. From inside the curb inlet the sight line to the car is low and partly blocked by the curb itself, the shooter would be in a concrete box in daylight beside a crowd, and no physical evidence has ever been found in it. It remains a hole in a curb that people find interesting because it exists.^{39, 3, 2}

There is a claim that circulates online that the Secret Service driver turned and fired. It comes from a poor-quality copy of the film where the sun catches Roy Kellerman’s hair in the passenger seat. Better transfers show it plainly. Bill Greer had both hands on the wheel and turned his head twice. This one is not a matter of interpretation. It is an artifact of a bad video copy.^{11, 12}

†Take the single most quoted case. A well-known journalist who had interviewed Ruby privately died in 1965 from a combination of alcohol and barbiturates, and the manner of death was recorded as undetermined. That is genuinely unresolved. It is also, in 1965, an extremely ordinary way for a person under strain to die. Unresolved and suspicious are not the same word, and the same certificate supports both readings. That is precisely why the lists work.^{40, 41}

†A widely shared clip appears to show an agent being waved back from the limousine at Love Field and stepping away with his arms out. It is real footage and it looks damning. Agents present described it as a routine repositioning as the cars got under way, and there is no order in any record. What the film shows is a man shrugging. What it means is not on the film.^{42, 43, 44}

†The most far-reaching claim about this case is that the Zapruder film itself was faked. Frames removed, images painted in, the whole thing rebuilt in a lab within days. It would have required matching several other films shot from other angles that nobody controlled. The film has real damage and real splices from handling. That is not the same as forgery, and nobody has ever shown the two other films disagreeing with it.^{45, 46, 12}

Almost every witness in this case has been called a liar by somebody. Very few of them were. People misremember direction under stress, compress time, and absorb details from what they read afterwards without knowing it. If you assume that everyone who contradicts your preferred version must be dishonest, you will end up with a version of that day held together entirely by accusations.^{47, 48}

Why famous names get pulled in

Right after the film came out I had dinner with Tommy Lee Jones in San Antonio. He had just played Clay Shaw. And what struck me was that he was exactly as fascinated as everybody else. Asking about the hobos in the rail yard, the ones later accounted for. The subject of Woody Harrelson’s father came up. A man who had been inside that story professionally still wanted to talk about the same three photographs the rest of us do.⁴⁹

‡In 1980, during a six-hour standoff with police, a convicted contract killer called Charles Harrelson said he had shot John Kennedy. He was high at the time. He retracted it as soon as he came down, and said he had been having fun with the police. He was later convicted of assassinating a federal judge in San Antonio. He is famous in this case for a confession he himself withdrew.^{50, 51}

‡The reason his name attached to Dealey Plaza is a photograph. Researchers compared him to one of the three men walked out of the rail yard and decided the face matched. That is the entire basis. In 1992 Dallas police records identified all three by name. Three men who had been sleeping rough and were released the next day. He was not among them. A resemblance in a grainy picture is not an identification.^{52, 53, 54}

There is a pattern worth naming. A grainy photograph, a face that could be anybody, and a name that carries. It happened to a shoe salesman standing in a doorway who spent his life proving he was himself. It happened to a killer who volunteered a confession and took it back. The photographs are real. What gets projected onto them is us. Wanting the cause to be as large as the effect.^{55, 56, 57}

What people alleged

It was a clear, warm day in Dallas. One man on Elm Street was holding an open black umbrella. He stands almost exactly where the first shot strikes, and for fifteen years he was the most suspicious figure in the plaza. Then he came forward. Louie Steven Witt told Congress it was a heckle, aimed at the President’s father. The only man in Dallas doing it.^{58, 59, 60}

†One witness claimed she watched a rifle carried up the grassy knoll before the motorcade arrived. Julia Ann Mercer said she was stuck in traffic on Elm and saw a man take a long case from a truck and carry it up toward the fence. Years later she said the affidavit in the record had been altered and the signature was not hers. Nobody has ever resolved that.^{61, 62, 34}

†A young soldier claimed a bullet went past his left ear on the grassy knoll. Gordon Arnold said he was on leave, standing on the slope with a movie camera, and that a shot came from behind him so close he hit the ground. He said two men took his film. He first told it publicly in 1978. No photograph has ever been found that shows him standing there.^{63, 64}

†A Dallas sheriff’s deputy said he watched a man run down the grassy knoll and get into a waiting station wagon. Roger Craig testified that a light-colored Nash Rambler picked up a man on Elm about fifteen minutes after the shooting. He later said it was Oswald, who could not drive. Other officers did not support it. Craig lost his job, and died in 1975.^{65, 66, 67}

The rifle found on the sixth floor was first announced to the press as a German Mauser. Deputy Seymour Weitzman helped find it and said so under oath. The weapon entered into evidence was an Italian Mannlicher-Carcano. Weitzman later said he had simply been wrong, in a dark room, in the first hour. The least interesting explanation, and the most likely one.^{68, 69, 70}

The witnesses who say it was changed

There is a claim in this book that deserves more than the single sentence it got. A woman said the affidavit with her name on it was not what she had said. That is not one witness with a grievance. There is a small group of people who stood in that plaza and afterwards said the same thing in different words: that they were leaned on, or that the paper filed under their name did not say what they said.⁷¹

†Julia Ann Mercer was sitting in stopped traffic on Elm Street an hour and a half before the motorcade, beside a green pickup truck half up on the curb. She said she watched a young man take a long case out of the back of it and carry it up the slope toward the fence. She reported it that day, to the sheriff’s office and to the Bureau. So this is not a story that surfaced years later. It is in the record from the first afternoon.^{72, 73}

†Years afterwards she looked at the document in the published record and said three things about it. That it did not say what she had said. That she had never sworn it in front of a notary, although a notary’s seal was on it. And that the signature was not hers. She also said that when she later saw Jack Ruby on television she recognized him as the man behind the wheel of that truck, which is not in the statement filed under her name at all.⁷⁴

†Most witness disputes in this case are about memory, and memory always loses. This one is not. She was not saying she remembered it differently. She was saying a specific physical document has a specific forged mark on it, which is the kind of claim that could in principle be settled by an examiner rather than argued about forever. As far as I can establish, nobody ever put it to one. That is the part I find hard.⁷¹

†The other one you have already met. Jean Hill, the schoolteacher in the red coat standing beside Mary Moorman, who ran toward the fence because she said she had seen a man up there. Her complaint was about pressure rather than paper. She said federal officers questioned her hard that weekend, that she was told what she had and had not seen, and that the transcript of her testimony did not match what she remembered saying. She never backed off any of it.^{75, 76}

But I am not going to hand you Jean Hill without the problem, because the problem is serious. Her account grew. What she said in 1963 and what she was saying decades later are not the same account, and the later versions contain things the earlier ones do not. That pattern is the single most common way an honest witness becomes useless, and it does not require anybody to be lying. It only requires sixty years, a great deal of attention, and being asked the same question a thousand times.⁷⁷

This is where the New Orleans district attorney comes in, because witnesses who said the official record had misused them were precisely what his investigation was built to collect. He gathered them, he believed them, and he put some of them in front of a jury. Jean Hill testified at his trial in 1969. He wrote about the woman and the truck as someone he found credible. And that is the honest extent of it: he took them seriously when the federal investigations had not.^{78, 79}

Which did them no good at all, and this is the cruel part. That prosecution collapsed. The jury took under an hour. And afterwards, every witness associated with it carried the collapse around with them, whatever they had personally seen. Being believed by the wrong investigation is worse for a witness than not being believed at all, because it gives everybody else a reason to stop listening that has nothing to do with what they saw.⁷¹

So here is where I stop. I do not think there was a program of rewriting witness statements, because if there had been you would expect it everywhere and you find it in a handful of complaints. What I do think is that a small number of people said something specific and checkable about documents bearing their names, and that in sixty years, through two federal investigations and a review board, nobody with the authority to settle it ever bothered. That is not a conspiracy. It is the same failure of ordinary work this whole book keeps running into.⁷¹

Things that were wrong

Almost everything in this case that once looked settled has been revised at least once. Not by cranks. By ordinary correction, as better evidence arrived. That is what a healthy record looks like, and it is worth saying out loud, because the moment a subject stops being revised is the moment it stops being history and becomes belief.^{80, 14, 81, 82}

For fifty years the bullet from the stretcher has been described as pristine, by people arguing both sides. It is not, and never was. It is flattened along one side and it lost a measurable amount of its weight. Whether that damage fits what the official account asks of it is a genuine argument. Whether the word pristine was ever accurate is not. A single adjective did decades of work that the object itself could not support.^{37, 38}

For fifteen years, the man holding an open umbrella on a clear day was the most suspicious figure in Dealey Plaza. Signal man, dart gun, accomplice. Then in 1978 he walked into a Congressional hearing, brought the umbrella with him, and explained that it was a political heckle aimed at the President’s father. Almost nobody in Dallas would have understood the reference. That is the whole answer, and it took fifteen years to arrive because nobody had asked him.⁸³

The three men pulled out of a railroad boxcar an hour after the shooting were photographed being walked across the plaza, and for thirty years they were whatever anybody needed them to be. Assassins, operatives, men in disguise. In 1992 Dallas police arrest records were released and all three were identified: they were exactly what the officers on the day had said they were. Three decades of argument ended by paperwork that had existed the whole time.^{52, 53}

One of the most repeated claims linking Oswald to organized crime is that he asked a Cuban exile shopkeeper whether he belonged to the Cosa Nostra. In 1978 a Congressional investigator went back to a witness who had been in that store, and got a duller answer: the words were almost certainly Casa Roca. The name of the store itself. A claim that ran for fifteen years appears to have been somebody mishearing a sign above a door.⁸⁴

In 1979 a Congressional committee concluded there had probably been a second gunman. That is the only official finding of conspiracy in the history of this case, and it rested almost entirely on one police radio recording analyzed as sound impulses. Within three years the National Academy of Sciences had taken that analysis apart. The finding still stands in the report. Almost nobody who works on this now believes the evidence it was built on.^{14, 85}

Now the half nobody expects. Some things dismissed for decades as paranoid turned out to be true. The CIA really had been trying to kill a foreign head of state, and really had not told the Warren Commission. The FBI really did keep files on the Commission’s critics. Agencies really did withhold material from the investigators they were supposedly assisting. People who said those things in 1970 were called cranks, and they were right.^{21, 14, 86}

The mysterious deaths lists are still circulating and they still work on people, because a list of names with unusual endings is genuinely striking. The question that dissolves them is boring and unanswerable in their favor: how many people were connected to this case at all? The answer is thousands, over sixty years. Some of them were always going to die strangely. A list without a denominator is not evidence. It is arithmetic performed on only one side.⁸⁷

There is one number underneath all of those lists and you have probably heard it. A British Sunday newspaper reported in 1967 that an actuary had calculated the odds against that many witnesses dying in three years at a hundred thousand trillion to one. It is quoted constantly. What is almost never quoted is what happened when a Congressional committee wrote to that newspaper eleven years later and asked how the figure was produced.⁸⁸

The newspaper’s legal manager wrote back and said the actuary had been asked the wrong question. He had been given a list of named individuals and asked the odds of those particular people dying, which is not the question at all, because you can always assemble a list after the fact. He described the original story as a careless journalistic mistake and said the paper did not intend to repeat it. That answer is printed in the committee’s own volumes and the number has been quoted ever since anyway.⁸⁹

†And then, because the argument has to cut both ways, two names that survive all of that. In June 1975 Sam Giancana was shot dead in his own basement days before he was due to testify to a Senate committee about the plots against Castro. The following year Johnny Roselli, who had already testified, was found in an oil drum in the sea off Florida while the House committee still wanted him back. Those two are not a list without a denominator. Those are two witnesses, killed, on schedule.^{90, 91}

Here is the only thing I would ask anyone to take from this. Hold what you believe about that afternoon the way you would hold a claim you expect to have to revise, because on the record of the last sixty years, you probably will. The people who have been most wrong about this are not the ones who changed their minds. They are the ones who decided early and then spent decades defending the decision rather than the evidence.⁹²

The telex that vanished

‡There is a document that a great many people believe existed, and if it did, everything in this book changes. The account comes from a young man who worked the night shift as a security clerk in the New Orleans field office of the F.B.I. In the small hours of the seventeenth of November 1963, he said, a teletype came in from headquarters. Five days before the motorcade.⁹³

‡He gave the wording, and the wording is why nobody can leave this alone. Threat to assassinate President Kennedy in Dallas, Texas, November twenty-two dash three, nineteen sixty-three. A militant revolutionary group may attempt to assassinate the President on his proposed trip. He said it went to four offices. Mobile, Memphis, New Orleans and Dallas. He said he telephoned the duty agent. And he said that about a week after the assassination the telex and the folder holding it were gone.⁹⁴

‡Be clear about the size of it. A genuine teletype like that would mean the Bureau knew the city, knew the dates, knew it was a group and not a man, and that the President rode through that plaza anyway. It would make every argument about angles and elevations a footnote. It is the single most consequential document anybody has ever claimed existed in this case, which is exactly why it has to be tested harder than anything else, not less.⁷¹

In the late seventies a Congressional committee went at it directly. Four field offices would have received that message, which means clerks, agents, supervisors, a duty log in each one. Fifteen years later, not one other person had ever come forward to say they saw it. The committee found no copy, no log entry, no corroboration from anybody, raised serious questions about the account itself, and concluded the allegation was unfounded.⁹⁵

So why does everybody know this story? Because it is perfect. A warning, ignored, then removed. And because it was dramatized, and a scene in a film reaches a hundred times as many people as a committee volume nobody opens. That is the whole mechanism by which a claim a Congressional inquiry rejected in 1979 is still, in 2026, one of the first things anybody mentions to me about the F.B.I.⁷¹

I do not think that teletype existed, and I want to be straight about why I included it anyway. One uncorroborated recollection, produced years later, contradicted by every record in four offices, is not evidence. It is a story. I have spent this book asking you to hold me to a standard, and the standard has to bite hardest on the things I would most enjoy believing. This is one of those, and it does not survive.^{71, 14}

The woman on the highway

‡Two days before Dallas, on a highway near Eunice, Louisiana, a woman was thrown out of a moving car and left at the roadside. A state police lieutenant picked her up. Her name was Melba Christine Marcades and she went by Rose Cheramie. She was injured, and she was coming off heroin, and on the drive she began to talk. She was taken to a state hospital, where she said the same things to the doctors and nurses.⁹⁶

‡The account is this. She had been traveling with two men, she said. They were going to Dallas to pick up a shipment of heroin, collect money, and kill the President. She is said to have named the trip, the city and the days. And she said she knew Jack Ruby, and had worked for him. If that is what she said on the twentieth of November, it is the only advance warning of the assassination anybody has ever produced.⁹⁶

‡Then the part that gives it weight. After the assassination the lieutenant who had picked her up telephoned the Dallas police to tell them about the woman in the hospital who had said it two days early. Dallas was not interested. He went back and interviewed her again, and he was still talking about it fifteen years later. Whatever the truth of it, that is a working policeman who acted at the time as though he believed what he had heard.⁹⁷

Here is the thing that decides it for me, and it is buried in the committee’s own volume. In 1967 both the lieutenant and the hospital doctor gave accounts of the episode to the New Orleans district attorney’s office. In those 1967 accounts there is no indication she said anything about the assassination before it happened. The advance warning appears in what they told Congressional investigators in 1978. Fifteen years after the event, and eleven years after they were first asked.^{96, 98}

She did not live long. In September 1965, on a road near Big Sandy in east Texas, she was struck by a car and killed. That fact does most of the work in how this story gets told, and I want to be careful with it. She was a heroin user living roughly, and people living that way die on roads. It is not evidence of anything on its own. It is also, undeniably, the second time a car nearly killed her and the second time it happened on a highway at night.⁹⁹

Put this next to the teletype from the last few minutes and you will see the same shape twice. A dramatic advance warning. One or two people who tell it. Nothing contemporaneous. And the crucial detail arriving years after the first time anybody asked. That is not proof either story is invented, and I cannot prove a negative. But when a memory improves with age, the improvement is the part to distrust, and both of these improved.⁷¹

The rest of them, including the silly ones

‡The strongest form is this. The United States had been trying to kill Castro for two years and he knew it, so he got there first. It is not a fringe position. Lyndon Johnson said something very like it privately, more than once, and he had the whole file. Against it is the arithmetic of consequence. If it had ever been shown, the pressure to invade Cuba would have been unstoppable, and the one man with most to lose from that was Castro.^{100, 25}

‡This is the big one, and it is the argument the famous film is built on. In October 1963 he approved a memorandum providing for a thousand advisers to come home by the end of the year. Six weeks later he was dead, and the war ran for another twelve years. The claim is that he was killed to stop the withdrawal. It is the most cited motive in this entire field and it deserves to be stated at full strength before anybody touches it.¹⁰¹

†Now read it rather than quote it. The withdrawal was conditional on the war going well, which it was not. He was on the record repeatedly that he did not intend to lose Vietnam. The memorandum his successor signed four days after Dallas reaffirmed the existing policy rather than reversing it. Historians have gone at this for fifty years without agreement. Nobody knows what he would have done, and not knowing is a very different thing from the certainty the theory needs.^{101, 102}

In January 1964 the Warren Commission held a closed session about a rumor that Oswald had been an FBI informant, on a payroll, with a number and a monthly figure. The transcript survives and it is remarkable, because you can hear them realizing what it would mean if it were true. They investigated it and found nothing to support it. What matters is not the rumor. It is that the men running the inquiry took it seriously enough to shut the doors.¹⁰³

Weeks before the assassination Oswald walked into the Dallas FBI office and left a note for the agent handling his file. After Oswald was shot, that agent was told by his superior to get rid of it. He tore it up and flushed it. None of this came out until 1975, twelve years later, and nobody now knows what the note said. That is not a theory anybody invented. It is a documented destruction of evidence by the Bureau, admitted by the man who did it.^{104, 14}

‡There is a figure in the Depository doorway in two films shot just after the shooting, and a body of research argues it is Oswald, which would put him downstairs and end the case. The problem is the same one that ends Badge Man. At that magnification you are looking at a handful of grains and a shape, and people who want it to be him see him. It would be the single most important image in the case if it resolved. It does not resolve.^{105, 106}

‡In June 1963 he signed an executive order about issuing silver certificates, and a whole theory says he was killed for challenging the Federal Reserve. Go and read the order. It was a technical delegation tied to the repeal of a silver act, and its practical effect was to help wind silver certificates down rather than replace the currency. The document that supposedly got him killed does the opposite of what the theory needs it to do.^{107, 108}

‡There is a theory that Israel arranged it over the reactor at Dimona. The real thing underneath it is that Kennedy did press hard for inspections and there was a genuine and documented diplomatic dispute. What there is not is a single document, source or piece of physical evidence connecting that dispute to Dealey Plaza. It is inference stretched across an entire ocean, and it circulates most often in company that tells you what it is really for.^{109, 110}

Somebody will ask why a book like this bothers with the driver turning round and firing, or a currency order, or a reactor. Here is why. A reference that only answers the respectable theories leaves every unrespectable one standing, and the unrespectable ones are the ones actually circulating. Every claim in here gets the same treatment. Strongest honest version, then measured against the record, then a plain verdict. Dismissal is not an argument, and it never convinced anybody of anything.¹¹¹

Twenty two hours after the shooting the Director of the FBI telephoned the new President, and the call was recorded. Johnson asked about the visit to the Soviet Embassy in Mexico City. Hoover said they had the tape and the photograph of the man who was there using Oswald’s name, that the picture and the voice did not correspond to Oswald, and that it appeared there was a second person. That is not a researcher’s claim. It is the head of the Bureau, to the President, on tape, the next morning.^{112, 113}

The one where the driver did it

There is one theory I have left alone that I should not have, because it is probably the most watched of all of them now. It says the man driving the car turned around and shot the President himself. It lives on video sites, it has tens of millions of views, and people arrive at it honestly, because at a glance the film does appear to show something in his hand and a movement at exactly the wrong moment.^{114, 71}

So here is what that is. On a poor copy of the film, in the front seat, there is a bright shape that appears near the driver as he turns his head. It is not in his hand and it is not a weapon. It is sunlight off the head of the man sitting beside him in the passenger seat, and it looks like an object because generational copying has smeared it into a hard-edged blob. On a good transfer from the original it stops being a shape and goes back to being a highlight.^{114, 71}

Which makes this the purest example in the whole case of something I keep running into. The theory is not built on the film. It is built on a copy of a copy of a copy, where the losses have invented an object that was never in front of the lens. Every version of this argument I have seen uses degraded footage, and the argument disappears at higher quality. That is not a conspiracy being covered up. That is a compression artifact being promoted.^{114, 71}

There is also a simpler objection that does not need any film at all. He was in the left front seat of a left-hand drive car, with a man beside him, two people on jump seats immediately behind, and the President behind those. To shoot backward over his right shoulder he would have to turn most of the way round, past a passenger, and fire through a car containing four other people, in the open, in front of hundreds of witnesses and several cameras. Nobody in that car described anything of the kind.^{114, 71}

And I mind this one more than the others, because there is a real failure sitting underneath it and the invented one crowds it out. He was a chauffeur before he was an agent. When the shots came he turned to look instead of accelerating, which cost seconds that may have mattered, and he carried that for the rest of his life. That is a genuine thing to examine. Accusing him of murder is not an escalation of it. It is a replacement for it.^{114, 71}

The confessions

There is a category of evidence in this case I have not addressed and it is the one people find most convincing. Confession. Roughly once a decade somebody says they did it, or drove for the people who did, or loved the man who did, or heard it from the man who did. They are usually old. It is often a deathbed. There is frequently a book. And not one of these accounts has ever survived contact with a document.⁷¹

‡The most detailed version comes from a woman who says she was Oswald’s lover in New Orleans in the summer of 1963, and that the two of them worked on a biological weapon intended for Castro. It is vivid, it is internally consistent, and she has told it for thirty years. The difficulty is that almost every checkable element in it either cannot be corroborated or matches material that was already published. Detail is not the same as evidence. A good memory and a good library produce the same texture.^{115, 116}

‡Then the shooters. A convicted killer in an Illinois prison who says he fired from behind the fence. A hitman who said it during a standoff and withdrew it as soon as he came down. A former intelligence officer whose family produced a deathbed account naming other people. Every one of them fails on the same thing: not a single document, photograph, transaction or third party has ever been produced to place any of them in that plaza.^{117, 118}

It is worth understanding why these land so hard. A confession does something no document does: it gives you a person, admitting a terrible thing, apparently against their own interest. Our instinct is that nobody would say that unless it were true. But people do, constantly, and the reasons are ordinary. Attention at the end of an unremarkable life. Money. Belief that has hardened into memory. Police forces get false confessions to ordinary murders every year, and this is not an ordinary murder.⁷¹

So here is the test, and there will be more of these after this book is printed. Does the account contain one checkable fact that was not already public when they said it. One receipt, one flight, one photograph, one person who is not a relative. That is the whole test. Every confession in this case so far has contained a great deal that was already in print and nothing at all that was not. When one of them contains something new, I will change my mind that afternoon.⁷¹

† disputed   ‡ theory   § reconstructed — everything unmarked is established. Superscript numbers key to the sources below. Corrections

Sources for this chapter

118 sources. Every claim above is keyed to one of them. The complete list for the whole book runs to 1,276.

1. Dealey Plaza drainage geometry and survey
2. storm-drain hypothesis literature
3. the reconstruction's terrain and curb model
4. Warren Report ch.6 (Oswald impersonation reports)
5. HSCA review of the sightings
6. Zapruder, Nix and Muchmore films
7. Beverly Oliver's claims and their critiques
8. Gary Mack and Jack White, 1982
9. photographic analysis of the Moorman print
10. HSCA hearings, 1978, including Louie Steven Witt's testimony
11. high-resolution Zapruder film transfers
12. HSCA photographic panel
13. the assassination literature, 1964 to the present
14. HSCA final report 1979
15. Warren Report ch.3
16. HSCA forensic pathology panel
17. the alteration literature and its critiques
18. Warren Report ch.5 (Ruby)
19. HSCA findings on Ruby's associations
20. HSCA final report, 1979 (agency disclosure findings)
21. Church Committee 1975-76
22. JFK Records Act releases
23. HSCA final report 1979, findings on organized crime
24. Warren Commission and Justice Department deportation record, 1961
25. Church Committee interim report 1975
26. Ed Reid, The Grim Reapers, 1969
27. HSCA final report 1979, section on Jack Ruby's associations
28. Warren Report, appendix XVI
29. Warren Report 1964
30. Warren Report ch.3 (FBI marksman trials)
31. HSCA firearms panel
32. CE 884 range figures
33. Warren Report ch.4 (USMC marksmanship records)
34. HSCA review
35. Zapruder film Z313-316
36. the physics literature on both sides
37. Warren Report ch.3 (CE 399 condition and weight)
38. HSCA forensic panel
39. Dealey Plaza drainage geometry
40. 'mysterious deaths' literature
41. statistical critiques of the death-list claims
42. Love Field departure footage
43. Secret Service statements on motorcade formation
44. Warren Report ch.8
45. Zapruder film handling and splice records
46. the Nix and Muchmore films as independent checks
47. Warren Commission and HSCA witness records
48. eyewitness memory research
49. personal, the narrator's own account
50. contemporaneous reporting of the 1980 standoff and Harrelson's retraction
51. United States v. Harrelson (murder of Judge John H. Wood Jr., 1979)
52. Dallas Police arrest records released 1992
53. rail-yard photographs, 22 Nov 1963
54. photographic comparison literature and its critiques
55. the Dealey Plaza photographic record
56. HSCA photographic panel findings
57. research on proportionality bias
58. Witt HSCA testimony, 1978
59. Zapruder film
60. Nix and Muchmore films
61. Mercer FBI/Sheriff statements, Nov 1963
62. Mercer's later interviews (contradictory)
63. Arnold interviews from 1978 onward
64. the plaza photographic record (Moorman, Willis, Betzner, Bond)
65. Craig WC testimony (6H260)
66. Craig's later accounts (contradictory)
67. Dallas County records
68. Weitzman WC testimony (7H105)
69. Warren Report ch.4 (CE 139)
70. DPD first-day press statements
71. the author's assessment
72. Dallas County Sheriff's Department statement of Julia Ann Mercer, 22 November 1963
73. FBI report of interview with Julia Ann Mercer, November 1963
74. later statements of Julia Ann Mercer concerning her 1963 affidavit
75. Warren Commission testimony of Jean Hill, March 1964
76. later public statements of Jean Hill
77. comparison of Jean Hill's 1963-64 statements with her later accounts
78. State of Louisiana v. Clay Shaw, trial record, 1969
79. Jim Garrison, On the Trail of the Assassins
80. Warren Report
81. ARRB final report 1998
82. post-2017 releases
83. Louie Steven Witt HSCA testimony, 1978
84. HSCA investigator's interview report, 14 April 1978, RIF 180-10106-10008
85. National Academy of Sciences review, 1982
86. RIF 124-10062-10456
87. 'mysterious deaths' literature and its statistical critiques
88. HSCA final report 1979, correspondence with The Sunday Times regarding the 1967 actuarial claim
89. HSCA final report 1979, reply from The Sunday Times to the committee, 1978
90. contemporaneous reporting of the deaths of Sam Giancana, June 1975, and John Roselli, 1976
91. Church Committee and HSCA witness schedules
92. the author's stated method
93. HSCA final report 1979, on the allegations of William S. Walter
94. HSCA final report 1979, volume on the FBI, the Walter allegations
95. HSCA final report 1979, findings on the alleged 17 November 1963 teletype
96. HSCA final report 1979, volume X, staff report on Rose Cheramie
97. HSCA final report 1979, volume X, on the account of Lieutenant Francis Fruge
98. New Orleans District Attorney's office interviews, 1967
99. Texas death record and contemporaneous reporting, September 1965
100. HSCA final report 1979, findings on Cuban government involvement
101. National Security Action Memorandum 263, 11 October 1963
102. National Security Action Memorandum 273, 26 November 1963
103. Warren Commission executive session transcript, 27 January 1964
104. House Judiciary and Senate testimony of James Hosty, 1975
105. Darnell and Wiegman film footage, 22 November 1963
106. HSCA photographic evidence panel, 1979
107. Executive Order 11110, 4 June 1963
108. Public Law 88-36, Silver Purchase Act repeal, 1963
109. State Department correspondence on Dimona inspections, 1963
110. no supporting record in the JFK Assassination Records Collection
111. personal, the author's statement of method
112. Telephone conversation between President Johnson and J. Edgar Hoover, 23 November 1963, 10:01am, LBJ Presidential Library recording
113. Hoover memorandum to the White House and the Chief of the Secret Service, 23 November 1963
114. the reconstruction
115. published claims of Judyth Vary Baker
116. responses of researchers on both sides of the case
117. published confession claims, 1980s to 2000s
118. HSCA and later investigative responses