To the editors:

In May of 1950, when I was a 24-year-old physics graduate student at Princeton, I accepted the invitations of my mentor John Wheeler and Wheeler’s friend Edward Teller to join an accelerated effort in Los Alamos to design an H-bomb. Wheeler had given me a soft sell. Teller came to town and gave me a hard sell. The force propelling the acceleration was the Soviet Union’s first successful A-bomb test less than a year earlier.

I applied for a Q clearance and it was granted in a few weeks. By late June I was at work in Los Alamos.

Adhering to the lab’s security requirements was not hard. Most challenging for me was the requirement that all of one’s work be recorded in bound notebooks. There were to be no notepads, no loose pieces of paper, and no wastebaskets containing half-baked thoughts on nuclear matters. I mostly adhered to this requirement. So now, on some shelf in Los Alamos, sit bound notebooks containing my thoughts and calculational efforts. Then, as now, I used a fountain pen.

Taking classified information out of the lab was not hard. Briefcases were not checked at the guard gate, nor were they examined at the airstrip. On the few trips east that I took during my year there, I only put secret information in my briefcase, never in checked luggage nor in parcels to be stowed under the seat ahead. When the stewardesses did not notice or care, I put the briefcase behind my lower legs next to the front of my seat—otherwise it was under my arm on one side of the seat. In either location, if the briefcase were touched while I was napping, I would wake.

Conversation at the lab was free and open. Only once did I come up against need-to-know. I was present in a small conference room where Hans Bethe, on one of his fairly frequent visits to the lab, was meeting with some of the lab officials. When the conversation turned to the analysis of air samples collected by US aircraft in order to learn more about Soviet atomic explosions, I was shooed from the room. I had no need to know. But that incident was the exception.

The next year, 1951–52, back at Project Matterhorn in Princeton, we had 24-hour guard service in the shack where we worked. Of course we talked outside of that building and even wrote things on pieces of paper. But we were never casual about secrets and did our best to shelter them. In January 1953, Wheeler lost a classified document on an overnight train trip to Washington. It was quite an important one, containing information on the use of lithium deuteride as a thermonuclear fuel. Despite herculean efforts by Amtrak and the FBI, it was never found. President Eisenhower, learning of the incident, was outraged. Wheeler was exceedingly embarrassed.

One of my early jobs at Matterhorn in the summer of 1951 was hiring young women—yes, they were all women and all young—to perform secretarial chores, carry out numerical calculations, and draw graphs. All were swiftly granted Q clearance. I spent part of the next year living in a rented room in Washington by day and running what we considered a powerful computer, the SEAC at the Bureau of Standards, by night. We used a program coded by John Toll, Wheeler, and me that followed the “burning” of the deuterium fuel in the forthcoming test of the first H-bomb, codenamed “Mike.” For this work I needed an assistant. I hired a young man to operate the computer and deal with its hexadecimal output. There was no time to consider seeking a security clearance for him, so he worked without a clearance and with no real idea of what we were calculating. I could only assure him that it had to do with national security and was very important. Perhaps John von Neumann could have unraveled our machine-language code. I do not think anyone else could have. In any event, we operated without safes, without guards, and without bound notebooks.

Sixty years later, I began work on a book that, when it was published in 2015, bore the title Building the H Bomb: A Personal History.¹ Its purpose, as stated in the preface, was to inform interested nonscientists about the history of the world’s first H-bomb, provide a memoir of a two-year slice of my life, and serve as a mini-textbook on nuclear physics.

As I dug up recollections and records, did fact-checking—with the help of Anthony Eames, a very helpful graduate student at George Washington University—and wrote, now with a laptop rather than a fountain pen, it never occurred to me that there might be anything in the book that would catch the attention of the nation’s guardians of secrets. Everything in the book, I was sure, was by that time in the public domain. Yet, in a conversation in early 2014 with the Department of Energy’s historian, Terry Fehner, who had supplied me with some facts, Fehner said, “You know, Ken, since you held a Q clearance and were dealing with secret information in the time period you are writing about, you really should submit your manuscript to DOE for a security review.”

“I hadn’t thought about that,” I replied. “How long would such a review take?”

“Probably about a month.”

“OK, I can afford that. I’ll submit it for a pro forma review.”

In June 2014, I did so, and in August I heard back: “Our team is quite taken with your manuscript,” was the message. “However, some concerns have been identified.” After some back and forth, I agreed to meet with these keepers of the secrets at their office in Germantown, Maryland, for a discussion.

On the appointed day, I showed up around 9:00 a.m., and was told that for me to participate in a discussion of my manuscript, I needed Q clearance. The necessary photographing, fingerprinting, and form-filling was completed expeditiously. By 11:00 I was again a cleared scientist, a status that would last several hours. Then I sat down in the office of Andrew Weston-Dawkes, the head of the security section. A couple of other people were present. One of them had in his hands a copy of my manuscript with a forest of yellow Post-it notes sticking out from its pages. I was alarmed.

The message to me at this meeting ran along the following lines:

We like your book and would be happy to issue it as a DOE report, classified secret, so that it could reach people qualified to read it. Also, we would be glad to provide you with an office here in which you could complete and polish the manuscript, and work with our publishing team. If you want to publish the book commercially, you will have to make many changes and delete much material.

I declined the kind offer and said I wanted to publish the book commercially. We agreed that I would be sent a detailed list of everything in the book that the DOE officials wanted removed or changed. We parted on good terms and I went home, no longer in possession of a Q clearance.

About two months later, in November 2014, I received the list. It contained 60 items. Some of them called for a change, but most called for a deletion. In total, I was told to delete about 5,000 words, or about 10 percent of the book. In my response, I wrote, “Were I to follow all—or even most—of your suggestions, it would destroy the book.”

I worked my way through the 60 items, to see if any had merit. I found none that did. In some cases I could cite unclassified sources—although I acknowledge the argument that by stating something that has appeared elsewhere, I am authoritatively confirming its truth. Some were descriptions of events, say, in 1945 or in 1960, of which I had no direct knowledge. One was a statement that the main cylinder of the “Mike” device was about 20 feet tall and about 7 feet in diameter. This does indeed sound like classified information, but the DOE officials had no objection to including in the book a photograph of this device next to people, a Jeep, and a forklift that provided a scale of size.

I sent back my rebuttal. After a little more back and forth, I said I would replace “in fact” by “reportedly” in two places. In January 2015, we agreed that we were at an impasse and that there was no basis for further negotiation.

Going ahead required that I feel safe and that the publisher, World Scientific, felt protected. By good fortune, a law firm in Washington, DC, after hearing my story in detail and examining the evidence, agreed to defend me pro bono if it came to that. That took care of my safety. Every publishing contract, of course, assures the publisher that the author alone is responsible for any errors, false or libelous statements, and so on. After some amendment to that clause of the contract to make clear that any breaches of national security were my responsibility, World Scientific agreed to go ahead. And go ahead we did. The book was published in March 2015.

A little before the publication date I contacted Dennis Overbye, a senior science writer at the New York Times whom I had met earlier, telling him that I didn’t have a story for him yet but wanted to alert him to the possibility of a story if the DOE came after me. He relayed my message to his colleague William Broad, who said I was wrong, that in fact I had a story already. Broad came down to Philadelphia to spend some time with me, hear my account, and look at the written evidence, including my exchanges with the DOE officials. His article appeared in the science section of the Times on March 23, 2015.²

Nothing happened. No reverberations. No communication. Just an uptick in sales.

What do I conclude from all of this? I continue to think that secrets are necessary and that people entrusted with them do, with rare exceptions, take very seriously their responsibility to safeguard those secrets. Yet secrets are like helium. They escape from almost any container. They can’t be bottled for long. It is foolish to try to keep secrets for many decades. Andrei Sakharov, in a 1990 memoir that covers his work on the first Soviet H-bomb, refers to the first idea, the second idea, and the third idea that made possible the Soviet Union’s successful H-bomb program.³ In a 2004 book about Sakharov, Gennady Gorelik identifies these ideas, in effect, declassifying them.⁴ So these ideas originated in the 1950s, were first publicly named in 1990, and first publicly explained in 2004. That is a very long shelf life indeed for a secret.

For the record, the first idea was to alternately layer fission and fusion fuel, the second idea was to use lithium deuteride as a fusion fuel, and the third idea was to use radiation implosion to compress the fusion fuel. Sakharov used the term “layer cake” for the first idea, and it was implemented. Teller had the same idea and called it an alarm clock. It was not implemented. All three ideas circulated in our Los Alamos group in 1951. Historians agree that they arose independently in the Soviet Union without the help of espionage.

What about the keepers of the secrets? I bear them no ill will. They are doing a job that has been defined for them. But secrets erode with time and they leak. There’s not much these servants of secrecy can do about it.

Kenneth Ford

Jeremy Bernstein replies:

Reading Ken Ford’s comments brought back some of my own encounters with clearance. As I noted in my essay about the Orion project,⁵ I spent the early parts of the summer of 1958 at the RAND Corporation in Los Angeles. The headquarters was set inside what looked like a small university campus in Santa Monica not far from the beach. There were two levels of security. You needed a pass to get into the building and a full Q clearance to get past the guard for the physics group. At the time, the group was working on matters related to the hydrogen bomb. No one had anything for me to do until one day I was handed a long list of figures and asked to add them up. I was told that it had something to do with high altitude testing with a possible connection to intercontinental missile defense. It was around this time that Freeman Dyson rescued me with an invitation to move to La Jolla to work on the Orion.

Since it was going to be powered by nuclear weapons, a Q clearance was needed to work on the Orion project. But, as far as I was concerned, this was an irrelevance since what I did with Dyson had nothing to do with these weapons. But it did affect Richard Feynman. He refused to become a consultant. He said that he could not guarantee that he would not talk about what he learned about the project because he always talked about anything that excited him. What I worked on with Dyson was opacity—the absorption of radiation by plasmas, and the like. Dyson had proven a remarkable theorem on the maximum possible opacity using quantum mechanics. My job was to compare it with data. I knew that RAND was working on opacity for things like uranium and that this work was classified. We stuck to light elements. I persuaded Dyson to visit RAND with me to present his theorem. The lecture became something of a farce because Dyson did not know the clearance status of all the people attending and was unable to say anything about Orion. Although we had confined our detailed work to light elements, which avoided all the issues of classification, Dyson’s derivation applied to all elements.⁶ If he had been allowed to present this material, the attendees could have used it to check the validity of their approximation methods and perhaps saved themselves from a good deal of grief.

This connection with RAND led to an even more farcical situation. A few years later, I took a job at the Brookhaven National Laboratory. There was some classified work being done at the lab. I never knew what this was, but it meant that there was some level of security. One day I got a phone call from security letting me know that a classified document had arrived from RAND and asking for my permission to open it. I replied that they could open it, but only if I could read it. I was sure that it had something to do with the high altitude nuclear tests. I was told that I was not authorized to read it because my Q clearance had lapsed. The document was sent back to RAND unopened. I was listed on the cover as one of the co-authors.