What I Think

Listen to this chapter, read by the author →

The way it was done

‡I should be honest that I am not looking at this the way most people look at it. Most of the argument is about capability. Could that rifle do it, could that man do it, could an agency have arranged it. I do not think capability is the interesting question, because what I am looking at is bigger than what the Service, the Bureau or the Agency could have organized between them. I am looking at how it was staged, and staging is a choice.¹

‡Look at the order of it. He goes into the city he was warned about, and the crowds are enormous and they are friendly, and this book has already shown you that. By the time the car reaches that plaza he has every reason to think he has won the room and the state with it. And that is the moment. Not the airport, not the hotel, not a quiet road. At the end of a triumph, in front of his wife and a few hundred people on their lunch hour.^{1, 2}

‡Here is what I cannot get past. If the object is to remove a man, there are a hundred quieter ways and every one of them is safer. Daylight, open car, cameras running, a wife beside him, an entire city watching. That is not the cheapest way to do it or the most deniable. It is the most public. Whoever chose it wanted it seen, and I do not have a gentler word for choosing that than sport.¹

In my twenties I had a friend from South America and we got onto this film. He found it funny that I was struggling with the idea of a coup in my own country. He said it had happened several times in his, and he named them, and he was not being dramatic about it. He was describing weather. That conversation changed the shape of the question for me, because the thing I could not accept was not the event. It was that it could happen here.³

The man on the bench in that film was not invented. X is based on Colonel Fletcher Prouty, who was Chief of Special Operations for the Joint Chiefs during that administration and spent years running the Air Force side of support for clandestine Agency operations. He advised the production personally. So the most quoted speech in the picture is a real Pentagon officer making his own argument, through an actor, about a job he actually held.^{4, 5}

†Here is the part of Prouty’s story that stays with me. He was not in Washington that November. He had been sent to Antarctica with a Congressional party, and he came back through Christchurch in New Zealand. The shooting happened at seven thirty in the morning New Zealand time, and before noon an extra edition was on the stands with a quarter of the front page given over to detailed biography of a man who had not yet been charged.^{6, 7}

†Measure it before you enjoy it. New Zealand was about four and a half hours into the story, not minutes. The paper took its overseas copy from the wire services on teleprinters, on trust, handled by sub-editors rather than reporters. And the defection had been in American newspapers in 1959, so a file existed to draw on. That covers most of the speed. It does not entirely cover the volume, and reasonable people have argued about the remainder for fifty years.^{8, 9}

‡But listen to where he chose to stand when he told it. Not Dallas. Not Langley. A newsstand in New Zealand, told by a man whose actual job was running Air Force support for Agency operations worldwide. Of all the places to locate the thing that convinced him, he picked one of the five countries. I think he was pointing, and I think he was pointing carefully, in the only way a man in his position could. That is a reading of emphasis. It is not a document, and I am not going to dress it up as one.¹

Think about what that arrangement actually is, structurally. Five governments, five parliaments, five sets of oversight, and every one of them can only see its own agency. There is no joint committee. No single archive. No court anywhere with jurisdiction over the whole of it. It is not a body with a headquarters you could subpoena. Call it a pivot, a point the five turn on, and remember that a pivot is the one part of a machine nothing else can get a grip on.^{10, 11}

And there is a rule that holds it shut, and it is not secret. In intelligence sharing, material you receive from another service is not yours to release. The originator decides. It is called the third party rule and every one of those five observes it. So the moment anything about this case passed between them, it acquired a veto held in another country, outside the reach of any American law a president could sign. That is not a theory about the alliance. It is how the alliance works on an ordinary Tuesday.^{12, 13}

‡Put those two facts beside the one I started with. I said I was looking for something that did not need deniability. This is something better than deniability. It is an arrangement where no single government can be made to open the file, because the file is always somebody else’s to release, and the somebody else is always abroad. I cannot show you it did anything. I can show you it is the only structure in this story that was built so that nobody would ever have to.¹

Here is the silence I keep running into. The Bureau delivered its report to the Warren Commission in the first two weeks of December 1963. A president had been shot on the twenty second of November. In about seventeen days the investigating agency had reached its conclusion, written it up in volumes and handed it over. Everything that came afterwards was checking that answer. Almost nothing was opening a new line, because by December there were no new lines open.¹⁴

†So ask what an international inquiry would have looked like and then ask whether it happened. Toll records across borders, systematically. Cable and telegram traffic. Mail covers. Passport and border movements for anybody of interest, in both directions, for months either side. Some of that was done on the nightclub owner after the fact and it found things. Almost none of it was done outward, because outward was not a direction the investigation was pointed in during the first two weeks.^{15, 16}

‡So I will say the whole of it rather than the safe half. I do not think that arrangement is a filing system between friendly governments. I think it is controlled, and I think it is the real chain of power in the world, sitting above the governments that believe they own it. Presidents arrive and leave. That does not. If you accept that, then a president who becomes a problem is a personnel matter, and Dealey Plaza is what a personnel matter looks like when the intention is that everybody sees it.¹

‡Now the honest part, and I would rather say it than have you say it for me. Everything up to here you can check. The agreement, the third party rule, the two documents, the absence of any allied inquiry. All of it has a number or a date. The sentence before this one has neither. Controlled by whom is a question I cannot answer, and a claim you cannot test is a claim that cannot be wrong, which is the oldest weakness there is. I hold it anyway. I am telling you it is the weakest thing I own.¹

‡Now catch me, because this is the move I have spent a whole book warning you about. I look at a clean scene and hear professionals. Somebody else looks at the same clean scene and hears that there was nothing to find. Both of us are reading an absence, and an absence will hold any shape you press into it. The difference between us is not the evidence. It is what we brought. I know which shape I press into it, and knowing that is the only honest thing I can offer you at the end.¹

‡So I stopped arguing about whether it was a coup. Governments are removed by force somewhere in the world most decades and nobody finds that unthinkable. What I cannot place is the manner. A coup wants a quiet transfer and a plausible story by morning. This wanted an audience. Sixty years later the thing people cannot put down is not that it happened. It is the taste it left, and the taste came from how it was done.¹

‡Prouty called it a coup and this book has already measured the part of his case that rests on Vietnam. What I take from him is the smaller word. Execution. Not a killing, an execution, which is a killing performed in front of people so that they understand something. Everything about the staging reads that way to me. Bring him into a hostile city. Let the crowds be enormous and friendly. Let him believe he has turned it. Then take his head off in front of his wife, in daylight, and leave every camera running.^{1, 17}

‡And once you are looking at it that way you cannot stop the next question. Malcolm X in 1965. King in 1968. The President’s own brother in 1968. Every one of them was changing what was possible, and every one of them was stopped in public. I am not telling you those are one file. I have no document that joins them and I am not going to pretend otherwise. I am telling you that I find it very hard to look at that decade and see four separate accidents.¹

‡Which brings me back to who. Organized crime kills quietly, because publicity is the one thing that costs it money. The Bureau builds files and ruins people slowly. The Agency works abroad with deniability and it is very good at it. All three could have worked together and none of them would have chosen a motorcade. What I am looking for is something that did not need deniability, and there are very few things in the world that have ever been able to say that.¹

The thesis, and the man

‡Let me put the whole thing in one paragraph so you can disagree with it early. The Bureau could not have executed this. The Service would have walked into them doing it, because walking into exactly that is the entire job. The Agency is not this ambitious at home and could not have held the lid down afterwards. But whoever did do it had a strategy that worked perfectly, including an arrest in a movie theater ninety minutes later that ran like it had been rehearsed. That is the shape I am trying to explain.¹

‡Take the Bureau seriously for a second before dismissing it. An operation like this needs weeks of positioning inside a protective perimeter, and the people who own that perimeter are the Secret Service, whose whole existence is noticing exactly that kind of movement. Two federal agencies in the same city, one of them watching for precisely what the other would have to be doing. Somebody would have walked into somebody. That is not loyalty talking, it is arithmetic.¹

‡And then there is the movie theater. Ninety minutes after the shooting, a store manager notices a man ducking a doorway, follows him, tells a box office to ring the police, and something like fifteen officers converge on a picture house in a suburb. Every step of that is explicable and this book has explained it. Put it end to end at speed and it still reads to me less like a lucky afternoon than like something that went the way somebody expected it to go.^{1, 18}

Here is the fact that made me stop treating him as a drifter. At eighteen he was a Marine radar operator at Atsugi in Japan, and Atsugi is where U-2 reconnaissance flights went out from. He held a clearance. He tracked aircraft whose existence was secret. Then he taught himself Russian well enough to live there, defected, and came back. Whatever else that is, it is not the biography of a simple man, and the lone drifter version needs him to be one.^{19, 20}

In the nineties I went up to the sixth floor with a friend who was a sniper in the United States Army. We stood at a window a few feet from the corner and looked down at Elm Street. I asked him what he thought. He did not hesitate and he was not being provocative. He said it was a very easy shot. A trained man, standing in the place, looking at the actual distance, told me the thing everybody argues about was not difficult.³

We came down and walked round to the top of the knoll and stood behind the stockade fence, which by then was a remnant of the one that had been there in 1963, weathered and patched and part of it still the same wood. You look over it at the road and the distance is nothing. A man could not miss from there. And I stood in that spot, at the age I was, having read about it my whole life, waiting to feel something conclusive.^{21, 22}

What came instead was the opposite. Standing in the two best places anybody has ever argued about, on the same afternoon, I understood that neither of them was going to tell me anything that mattered. Both were possible. That was the whole finding. And whichever it was, the man in the car was dead either way, so the mechanics could not change a single thing about the outcome I actually cared about.^{21, 22}

I want to be careful here, because I then spent years measuring that plaza and I do not think that was wasted. Measurement is very good at one job: telling you what could not have happened. It closes doors. But it never opens the one you want. No paraffin test, no bullet path, no angle taken off a survey table has ever told anybody why a president was killed, and I had been treating the hard part as though it were the interesting part.^{21, 22}

What tells you more is the other set of questions entirely. Who was actually capable of doing this. Who would have had to be, and who would have noticed. And then, far more than any of that, how everyone behaved afterwards. What was filed and what was not. What was destroyed in the following two weeks and what was destroyed in 1995. Who asked for the file and who quietly did not. Conduct is evidence, and there is sixty years of it.^{21, 22}

That afternoon is the reason this book is shaped the way it is. It opens with a street measured to the inch and it does not end there, because the street was never going to finish the argument. It ends in a statute, and in four filing systems belonging to governments that are not ours. I built the plaza to find out what the ground permits. Everything I actually believe came from watching how people handled the paper afterwards.^{21, 22}

‡I have thought about that for thirty years and I am going to be straight about what it costs me. If the shot was easy, then the whole marksmanship argument goes, and a great deal of conspiracy writing rests on it being hard. It is the weakest leg of the case I am making, and the man who kicked it out was standing next to me, qualified, with no stake in any of it. What survives is not that it took a marvel of shooting. It is everything about how it was arranged and what happened afterwards.¹

‡So let me be exact about what I think, because could have and did are different words and most of this argument runs them together. Yes, he could have made that shot. My friend at the window settled that for me thirty years ago and I am not going to pretend otherwise to protect a position. Capability was never my objection. I simply do not think he was the man who fired, and nothing about the shot being easy makes him the one who took it.¹

‡So I think he was an asset, and I think that is why he walked into a Bureau office and left a note for an agent weeks before. Men do not do that casually. This book already gives you two things that sit under it. A commission that closed its doors over a rumor he was on a payroll, and an agent ordered to destroy the note he left. Neither proves he was anybody’s. Both are exactly what you would expect to find lying around if he was.^{1, 23, 24}

‡If he was in that building on purpose rather than by the accident of a job he got through a neighbor, the role I can make sense of is not shooter. It is spotter. The one who watches, calls the car, and is not looking through the sight when it happens. And my objection to him firing is not about marksmanship, because my friend at that window settled marksmanship for me. It is about the clock.¹

‡Run it as a sequence. Fire three times from a corner window. Wipe down and hide a rifle between stacks of boxes forty-five feet away. Cross the floor, take a wooden staircase four flights down past a building full of people, and be standing in a second-floor lunchroom, not out of breath, holding a soft drink, when a policeman comes through the door with his gun up. About ninety seconds. Every part of that has been tested and argued and it is always close. It has never once been comfortable.^{25, 26}

‡On top of the clock there is a shape in a doorway. Two newsreel cameramen filmed the front of that building within a minute of the shots, and standing back in the shadow of the entrance is somebody who was never identified. I deal with him properly later in this book, including the reasons he is probably somebody else. He is not evidence. He is the reason I cannot make the ninety seconds sit still.²⁷

‡Now the thing that argues hardest against me, and it is physical. His palm print is on the underside of that rifle barrel, in a position you can only reach with the weapon partly taken apart. Spotters do not strip the rifle. I can say his hands were on that weapon at some point without his eye ever being behind the sight, and that is a real distinction. It is also the kind of distinction people make when they are defending a position rather than testing one, and I know it.^{1, 28}

‡So I will go further than asset. I do not think Lee Harvey Oswald was only a man. I think by 1963 the name was a working identity, built up over years and wearable by more than one person, which is a thing intelligence services actually construct and have a word for. A legend. If you accept that, the sightings in places he was not stop being bad memory, and the man at that embassy stops being a mystery, and the arrest of a single confused person in a movie theater stops being the end of anything.¹

‡And now the size of what I have actually got. Hoover’s call establishes one impersonation, in one city, in one season. It does not establish years of it, or a second man living the name, or a program. There is a long distance between somebody used Oswald’s name at an embassy and Oswald was a suit that people took turns wearing, and everything I just said lives in that distance. I think the record points that way. I am telling you it does not arrive.^{1, 29}

‡And notice what that call means about the file. Twenty two hours after the shooting the Director already had the Mexico City tape and photograph in front of him, compared, and a conclusion formed. Some of that is simply the Bureau moving fast on the biggest case it would ever have. But material of that kind does not assemble overnight from nothing. Somebody had been collecting on that name well before anybody in Dallas had heard it.²⁹

The shape of the six seconds

‡Here is something about the sequence I cannot let go of. The earlier hit did not kill him. He was struck, he was reacting, his hands came up, and the car was still moving through it. Then there is a pause, and then the last one, and the last one is different in kind. Not another attempt at the same thing. A conclusion. Whatever you think about who fired, the shape of those six seconds is a wounding followed by an ending, and they do not feel like the same act.^{1, 30}

‡Read that alongside everything else about how it was staged and you get the thing I keep arriving at. The first shots get the car, the crowd and the cameras looking. The last one is the sentence. That is what an execution is for, as opposed to a killing: the earlier part gets the attention and the final part is the statement. I am describing what the sequence looks like to me. I am not claiming to know it was designed that way.¹

†Now the explanation that needs no design at all, and I think it is the more likely one. A shooter fires at a moving target and takes what presents. The first shot is rushed at a car that is turning. The second finds the body, which is what anybody aims at. And by the third the car has slowed almost to nothing and the target is clear and close, so the shot lands where it lands. Everything I just called a sentence is also just a man getting better as the range improved. I hold my reading loosely.^{1, 31}

What the arrangement was

The arrangement did not appear from nowhere. It grew out of the wartime codebreaking partnership between Bletchley Park and the American cryptanalysts, formalized in 1943 while the war was still running. In 1946 that became a permanent peacetime agreement between Britain and the United States, and Canada, Australia and New Zealand were brought in over the following decade. Five countries, one arrangement, built by people who had spent the war reading somebody else’s mail and saw no reason to stop.^{32, 33}

By November 1963 this was the largest intelligence collection enterprise that had ever existed. Listening stations across five continents. Thousands of people whose entire working life was intercepting, decrypting and translating the diplomatic and military traffic of the Soviet bloc and its friends. The Cuban missile crisis had happened thirteen months earlier and Cuba was among the most heavily covered targets on earth. That is the machine that was running that fall.^{34, 35}

You do not have to take my word for the coverage, because it is in chapter nine. Six months after Dallas a device inside the Cuban Embassy in London, run by British domestic security and passed onward through British foreign intelligence, was picking up what Cuban diplomats said to each other in a private room. That is what routine coverage of Cuba looked like in that period, in an allied capital, shared onward to Washington.³⁶

‡So hold the two halves together. An arrangement covering Cuban and Soviet traffic worldwide, at the height of its powers, thirteen months after the missiles. And a man who walked into both the Cuban and the Soviet embassies in Mexico City seven weeks before Dallas, asking for a visa, and was refused. That is precisely the kind of traffic the whole apparatus existed to hear. The proposition on the table is that it heard nothing worth writing down, and that in sixty years no partner has ever been asked whether it did.^{37, 38, 35}

You do not have to wonder whether agencies ran things past presidents, because the Senate went and looked. In 1975 the Church Committee documented plots against foreign leaders across several administrations and found that deniability had been deliberately built into them, so that responsibility could not be traced upward. The Committee then said it could not categorically establish whether any President had approved. That is the United States Senate reporting that it could not tell who authorized assassinations carried out in America’s name.³⁹

A telephone call in England

On the twenty second of November an anonymous caller rang the senior reporter of the Cambridge News, in England, and said he should telephone the American Embassy in London for some big news. Then he hung up. The call was logged at five past six in the evening, Greenwich time. The shots in Dallas were fired at six thirty in the evening, Greenwich. Somebody in England told a provincial newspaper to expect news out of America twenty five minutes before there was any.⁴⁰

Now here is where I have to correct something I said. MI5 did look into it. They timed the call, they identified the reporter, and they reported that he was known to them as sound and loyal with no security record. That is a Five Eyes service running an inquiry into the assassination and writing an assessment. So the silence is not total. It is one memorandum, and we only have it because the CIA passed it to the FBI and the Americans released it sixty years later.⁴¹

‡So the one piece of allied investigation anybody can read reached the public through Washington rather than London, sixty years after it was written, because an American statute compelled an American agency to open an American file. Britain has still never released a thing about it. I cannot tell you what else MI5 did or did not do that week. Neither can you, and neither can any British citizen, and that is the whole of my complaint.¹

And notice how the one exception reached us. Not from London. It came out because an American statute compelled an American agency to open an American file, and the memorandum happened to be sitting in it. Britain has never released anything about this of its own accord, and neither has Canada, Australia or New Zealand. Everything anybody knows about what the allies did that week, we know because Washington was made to tell us and they were not.^{42, 43}

Let me read the document again, slowly, because I have been leaning on it. A CIA counterintelligence memorandum of the twenty-sixth of November 1963 reports what the British Security Service had passed on. That at about five past six in the evening, London time, an anonymous caller telephoned the senior reporter of the Cambridge News and told him to ring the American Embassy in London for some big news. That the caller then rang off without saying anything else. That the President was shot twenty-five minutes later.⁴⁴

Now here is the whole of the British contribution, and it is shorter than people assume. They established the time. They established that the reporter had never had a call like it before. And they said he was known to them, with no security record, a man they regarded as reliable. That is it. That is everything the released record shows the Security Service of the United Kingdom did about apparent advance knowledge of the assassination of the American president.⁴⁵

Before I make anything of it, the objections, and they are not trivial. The call was reported after the assassination, not before, so the timing rests on a man’s recollection of an ordinary evening that had not yet become memorable. Newspapers receive anonymous calls constantly and most are nothing. Big news is not a description of anything. And in 1963 there was no realistic way to trace an incoming call to a newspaper switchboard, so nobody could ever have found the caller even if they had tried very hard.²²

‡And having said all that, one question survives it and I have never had a good answer. Why Cambridge. If somebody wanted to leak, you leak to Fleet Street or a wire service. If somebody wanted to boast, you tell a person. Nobody chooses the senior reporter of a provincial English paper, by telephone, and gives him twenty-five minutes and no detail. It is the wrong target for every ordinary motive I can think of, and being wrong for all the ordinary explanations is precisely what makes something worth looking at.²²

And I have to be careful about what it would establish even if every word of it is exact. It would show that a person in England had reason to expect news from America that evening. It would not show they knew what the news was, or who was going to do it, or that they were connected to anybody in Dallas. Advance knowledge of a thing and involvement in a thing are different, and the distance between them is where most of the bad writing about this case lives.²²

‡So my complaint is not really about the call. It is about what was done with it. A security service is told that somebody on its own territory appeared to know, before it happened, that something enormous was coming out of America. And the surviving product of its inquiry is a paragraph confirming the time and a character reference for a journalist. No trace attempt on record. No canvass. No follow-up anybody has ever seen. That is not an investigation. That is a note.^{22, 45}

And here is the thing that keeps this at the center of my chapter. I do not know whether that paragraph is the whole inquiry or the visible corner of one. If the Security Service ran a serious operation on that call, the record of it is in London and no British law obliges anybody to say. If it did nothing beyond writing a note, that is remarkable in a different way. Sixty-three years on, a British citizen cannot find out which of those two things happened in their own country.²²

Everybody asks the wrong who

I have made the journey to Arlington and stood at the grave. It is sobering in a way I was not ready for. There is a flame there that has been burning since his widow lit it at the funeral, and like almost everything else in this story it is smaller than the photographs suggest. You stand on a hillside above a city he ran for a thousand days, and the whole argument goes quiet for a minute, and then it comes back.^{3, 46}

‡Sixty years of this has been spent on one question. Who did it. Who was at the window, who was behind the fence, whose finger was on which trigger. I think that is the wrong who. A trigger is the cheapest part of an operation like this and always has been. The question that actually sorts the field is not who fired. It is who is powerful enough to arrange something on this scale and then live comfortably with the consequences for sixty years.¹

‡The two favorite answers do not survive that question. The CIA is extraordinary at what it does and this book has spent chapters showing it, and what it does is operate inside somebody else’s country with deniability. The Bureau in 1963 was a filing system with a director’s ego attached to it, and it could not keep a routing slip in the right drawer. I am not saying either is innocent of everything. I am saying neither is the size of thing I am looking for.¹

‡Here is how I get there, so you can take it apart. Look at what happened afterwards rather than what happened in six seconds. A record sealed for decades. Files still coming out sixty years later. Two federal investigations that were not told what their own government had been doing. Whatever moved here was able to shape what an entire state could know about its own president, for longer than most people are alive. That is not a capability a small group has.¹

‡So I will name it rather than gesture at it. What I think this points at is the intelligence relationship between the English speaking states, the thing now called Five Eyes. Not one agency in one country, but the standing arrangement between them, which was already fifteen years old in 1963 and is still running. I cannot find another actor with the reach to do it, the reach to hold it, and no single government able to open it up afterwards.¹

‡The instinct underneath all of it is simple and I will not dress it up. Things like this do not just happen. A head of state is not killed in daylight in front of hundreds of people, in a country with the resources of that one, and then have the answer stay unsettled for sixty years, by accident. Something either arranged that or was content to let it stand. Both of those require a great deal of power, and only one of them requires a gunman.¹

‡Everybody asks what he was about to do that got him killed. I have come to think the better question is what stopped happening. He died during a thaw and this book has already shown you that. Test ban signed in August. A telephone line strung to Moscow so two men could speak before anybody launched. A speech in June about breathing the same air. All of it going one direction. Sixty years later the confrontation he was trying to end is still running, and nothing since has come as close to ending it.^{1, 47, 48}

The law that stopped at the water

It is worth saying what that is, because it is not a rumor. In 1946 Britain and the United States signed a signals intelligence agreement, and Canada, Australia and New Zealand were brought in over the following decade. Five countries, sharing what they intercept, under a classification marking that lists all five. It is real, it is public, it predates the Kennedy administration by fifteen years, and it has outlived every person in this book.⁴⁹

I went and read what the arrangement commits them to, because it matters. The 1946 agreement, declassified by both governments in 2010, is about signals intelligence, and the parties undertake to exchange the products of collection. Intercepted traffic. Traffic analysis. Cryptanalysis. Decryption and translation. It is not a friendship. It is a standing obligation to hand over what you hear, and it was seventeen years old and fully running by November 1963.³²

You do not have to take my word for whether they were listening, because it is in this book already. Six months after Dallas a device inside the Cuban Embassy in London, run by British domestic security and shared onward through British foreign intelligence, picked up two men discussing the assassination. The CIA station received it. The FBI cabled it to Washington. Four services, three countries, one overheard conversation, filed under the assassination case number, with a record number you can look up.³⁶

Now hold that next to the other document. In 1968 the Director of the CIA wrote to the head of Australian intelligence, another member of the same arrangement, to thank him for recommending against declassifying a Warren Commission paper. That is the only glimpse the released record gives us of these services talking to each other about what the public should see of this case. Not sharing something. Keeping something closed, and thanking a partner for the reasons.¹³

And here is the part that made me stop. That relationship is in this book already, from the 2025 files, before I ever got to this chapter. British intelligence bugging the Cuban Embassy in London and passing the product to the Bureau. A CIA station in Paris handing the Bureau a letter. The Bureau asking Madrid about it. A cable where the other Government agency is decoded in writing. Nobody had to theorise any of that. It is in the released record with numbers on it.⁵⁰

I said this book already had that relationship in it, and I want to be exact about how much. Fifteen documents from the release are read closely in these pages. Two of them involve a Five Eyes partner. One is the bug in London. One is the letter to Australia. That is eight passages of this book resting on two pieces of paper, and you deserve to know that before you weigh anything I have said. Mentions are not evidence. Documents are.⁵¹

There is a count that cuts the other way and it would be dishonest to leave it out. Five of those fifteen documents involve liaison with services that are not in that arrangement at all. Spain. France. Switzerland. Austria. On the raw numbers, the released record shows this case leaning on non allied services considerably more than on allied ones. You can read that as the ordinary business of an investigation, or you can read it as the allied material being the part that never came out. I know which reading I prefer, and preferring it is not the same as being right.⁵²

Here is the thing about the great releases everybody celebrates. The Records Act reaches United States government records, and state and local police records, held in United States hands. That is its whole territory. It has no reach into any other country’s files, and it never did. So when a president signs an order and the files come out, what has actually opened is the American share of an arrangement that had five members.⁵³

Congress did notice the problem. What it wrote is section ten, subsection b, paragraph two, and it is a sense of Congress rather than a command. The Secretary of State should contact foreign governments and ask them to disclose what they hold. Should. Not shall. No deadline, no reporting requirement, no penalty, no review board. Everything I have said in this chapter about allied files being closed follows from that one word, chosen in 1992, and so far as I can establish the provision has never once been used.⁵⁴

Now read the whole sentence, because the shape of it is the argument. Congress named the adversary. It said the Secretary should contact the Government of Russia and seek all records of the former Soviet Union, and then it named the K.G.B. and it named the G.R.U., in full, in the statute. Then came the allies, and for them Congress wrote this: and contact any other foreign government that may hold information. Named enemies. Generic friends. In the same sentence.⁵⁴

Here is the negative evidence, and it needs stating carefully because there is a fair objection to it. Intelligence services do not publish internal assessments. Nobody should expect them to, and an absence of published product proves nothing on its own. So my point is not that those four governments have said nothing. It is that nothing has ever required them to look. No British equivalent of the Records Act. No Canadian one, no Australian one, no New Zealand one. No mechanism, no deadline, no obligation of any kind has ever been pointed at them.^{42, 55}

So I asked them

Everything I have told you about the allied files has been an argument from silence, and I have been uncomfortable with that for a hundred pages. No partner government has ever been asked this question in a form it has to answer. So rather than write another chapter reasoning about what they might be holding, I wrote to four of them and asked.⁵⁶

The question is narrow on purpose, because a broad one gets refused on scope and teaches you nothing. It has two halves. Does the department hold any record created between the day of the assassination and the end of 1964 relating to it. And has any such record been withheld, in whole or in part, at the request of or by agreement with the government of the United States. The second half is the entire point of the exercise.⁵⁶

And here is why I expect to lose and do not mind. If a government refuses that second half by citing a rule about information obtained in confidence from another government, then it has just told me, in writing, over a signature, on a dated letter, that it holds American material it is not free to release. That is not a consolation prize. That is better than a document. A release gives you one page. A refusal on those grounds gives you the arrangement.⁵⁶

They are not equally answerable and it is worth knowing which is which. Britain has an absolute exemption for anything touching its security bodies, so Britain will almost certainly say nothing, and may refuse even to confirm or deny. Canada has a specific clause for information received in confidence from a foreign government, which is the exact clause I am fishing for. Australia must examine records of that age and give a written decision with reasons, which can be appealed. New Zealand has no blanket exemption at all and refusals go to an ombudsman who overturns them.⁵⁶

Which is what makes this worth the postage. If all four refuse on foreign-government grounds, the argument in this book is proved in their own words rather than mine. If one of them releases something, it is a document nobody has read in sixty years. And if they contradict each other, four allies giving four different answers about the same weekend is itself a finding. There is no result here that leaves the page blank.⁵⁶

What I cannot do is tell you how it came out. These went out while this book was being finished, and the response times run from twenty working days to several months, with appeals beyond that. So this is the one thread in the book I have deliberately left hanging, and I would rather hand you a live question than a tidy paragraph. Whatever comes back will be published in full, refusals included, at the handle on the last page.⁵⁶

The part worth sitting with is the cost. Four envelopes. One of them needed five dollars. Sixty years, two thousand books, two federal investigations and a review board with subpoena power, and as far as I can establish nobody had ever simply written to Canberra and asked. I am not claiming that as brilliance. I am reporting it as a gap, and it is a gap the size of a postage stamp.⁵⁶

What I think the motive was

‡For years I thought the money answered it. The killing was never the payday, I would say. Vietnam was. Six seconds in a plaza is the film and the twelve years after are the merchandising, and merchandising is where the money has always been. I still think the benefit is worth counting. But it stopped satisfying me, because the people who got rich on that war and the people I am pointing at are not the same people.¹

‡What I have come round to is smaller and worse. I think it was personal, and I think the person it was personal about may not be the one everybody assumes. Consider Fidel Castro not as an ideology but as a man with utility. An island, ninety miles out, outside American law, useful to a great many people for a great many purposes. Look at what a cartel is worth today and then imagine a whole country available to somebody.²²

‡And here is why that would put a president in the way. He was trying to have Castro killed, and then, in the last months, he was trying to talk to him. Either one disturbs an arrangement. Kill the man and the utility ends. Make peace with him and the island stops being outside the law. A president who is unpredictable about something valuable is a worse problem than one who is simply hostile to it, and he had become exactly that.²²

‡Now stop me, because I have just done the thing this book was written against. That is a motive built out of plausibility and nothing else. I cannot give you a number for it. I cannot name a person who profited from that island in a way that reaches Elm Street. It is an intuition about how power actually works, dressed up in a paragraph, and you should treat it exactly as well as it deserves, which is not very.²²

Where I stop

‡There are only two answers and I would take either. The first is that they genuinely held nothing. The second is that something was held and no law that exists can compel it. I used to say I did not know which. I have stopped saying that, because the first one has a problem I cannot get past.¹

‡Here is the problem with nothing. We are not looking at a blank wall. We are looking at prints on it. A British device inside the Cuban Embassy in London, picking up two men discussing the assassination, passed onward to Washington. A CIA director thanking Australian intelligence for advising against a release. An FBI airtel from Bern decoding the other Government agency in writing. Those are smudges left by something moving through the room. If the answer is that nothing was there, then what left the marks?^{22, 57}

‡So I am going to leave the question open, because I cannot close it, and I am not going to pretend the two answers are equally weighted. Held nothing requires me to believe that the largest listening arrangement ever built, at the height of its powers, covering Cuba as a priority target, produced not one assessable document about the murder of the senior partner’s own president. I can believe it. I have to work at it. The other answer requires no effort at all.²²

‡Now the part I have to say, because the whole book falls over if I do not. Showing that this relationship was operating on this case is not showing that it killed anybody. Those four documents prove reach and interest. They do not prove intent and they do not prove an act. I am telling you my conclusion goes further than my evidence does. If I let those two things blur here, after twelve chapters of insisting they are different, I would deserve everything a critic did to me.¹

‡I said at the front that everything in here would be marked established, disputed or theory, and that I would never let one sound like another. So here it is, marked. This chapter is theory. It is the only part of this book I cannot hand you a record number for. Everything else you can check without me and I would rather you did. If you finish this and disagree with this last chapter but trust the other twelve, then the book worked exactly as intended.¹

I have spent this book telling you what would falsify other people’s claims, so here is mine. If those four governments opened their files under a real statute, with a board and a deadline, and there was genuinely nothing on this case, I would withdraw this chapter and say so in the next edition. That is the test. It is not a hard test. Nobody has ever had to sit it.²²

‡And this is the thing I keep coming back to, underneath all of it. Somebody murdered the head of state of their closest ally. Not a rival. The senior partner in an arrangement they had built together and depended on. The response should have been the hunt of all hunts. Every station, every source, every line, turned on and pointed at it, out of loyalty and out of pure self-interest, because if it can happen to him it can happen to anyone. That is what I would expect from a friend. I cannot find any sign that it happened.²²

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

Sources for this chapter

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

1. personal, the narrator's own opinion
2. crowd accounts cited at M0221 and M0611
3. personal, the narrator's own account
4. L. Fletcher Prouty, Chief of Special Operations for the Joint Chiefs of Staff, 1955 to 1964
5. credited creative advisor, JFK, 1990 to 1991
6. L. Fletcher Prouty, statements to the Assassination Records Review Board
7. The Christchurch Star, 23 November 1963
8. The Christchurch Star, 23 November 1963, wire service sourcing
9. contemporary American press coverage of the 1959 defection
10. UKUSA Agreement, 5 March 1946, declassified 25 June 2010
11. absence of any joint parliamentary or judicial oversight mechanism across the five parties
12. third party rule in intelligence liaison practice
13. RIF 104-10009-10222, released 18 March 2025, National Archives JFK Collection
14. Federal Bureau of Investigation summary report to the President's Commission, December 1963
15. Warren Report 1964
16. HSCA final report 1979, findings on the adequacy of the original investigation
17. L. Fletcher Prouty, JFK: The CIA, Vietnam, and the Plot to Assassinate John F. Kennedy, 1992
18. arrest sequence established at M0505 and M0506
19. United States Marine Corps service record of Lee Harvey Oswald
20. Warren Report, appendix XIII
21. the author's recollection
22. the author's assessment
23. Warren Commission executive session transcript, 27 January 1964
24. testimony of James Hosty, 1975
25. Warren Commission Report, chapter IV, timing reconstructions
26. Warren Commission testimony of Marrion Baker and Roy Truly
27. Wiegman and Darnell films, 22 November 1963
28. palm print evidence established at M0116
29. Telephone conversation between President Johnson and J. Edgar Hoover, 23 November 1963
30. frame sequence established at M0212 to M0214
31. car speed established at M0639 and M0215
32. UKUSA Agreement, 5 March 1946, declassified jointly by the National Archives and the NSA, 25 June 2010
33. BRUSA Agreement, 1943
34. Church Committee interim report 1975
35. UKUSA Agreement, declassified 2010
36. RIF 124-10058-10420, released 18 March 2025, National Archives JFK Collection
37. Warren Report, chapter VI
38. RIF 124-10058-10420
39. Church Committee interim report, Alleged Assassination Plots Involving Foreign Leaders, 1975
40. Memorandum from James Angleton, CIA Counterintelligence, to the Director of the FBI, 26 November 1963, JFK Assassination Records Collection
41. Memorandum from James Angleton, CIA Counterintelligence, to the Director of the FBI, 26 November 1963
42. President John F. Kennedy Assassination Records Collection Act of 1992
43. Memorandum from James Angleton, CIA Counterintelligence, 26 November 1963
44. CIA counterintelligence memorandum, 26 November 1963, reporting British Security Service information
45. CIA counterintelligence memorandum, 26 November 1963
46. Arlington National Cemetery, gravesite of John F. Kennedy, eternal flame lit 25 November 1963
47. Limited Test Ban Treaty, August 1963
48. thaw material established at M0610
49. UKUSA Agreement, 1946, released by the National Archives and GCHQ in 2010
50. RIF documents cited at M0380 to M0385, M0390 to M0394 and M0450
51. RIF 104-10009-10222 and RIF 124-10058-10420, released 18 March 2025, National Archives JFK Collection
52. RIF 124-10115-10027, RIF 124-10150-10104, RIF 124-10062-10456, RIF 104-10069-10112, RIF 124-10058-10420
53. President John F. Kennedy Assassination Records Collection Act of 1992, definitions
54. Public Law 102-526, section 10(b)(2), 26 October 1992
55. absence of equivalent statute or public inquiry in the United Kingdom, Canada, Australia or New Zealand
56. the author's correspondence with allied governments, 2026
57. RIF material discussed earlier in this book