To the Grave: Secrets, Sins, and Nuclear Insecurity

By Joshua RoebkeMay 19, 2021

To the Grave: Secrets, Sins, and Nuclear Insecurity

Restricted Data: The History of Nuclear Secrecy in the United States by Alex Wellerstein

SECRECY AND SECURITY do not bloom from the same Latin root, but as ideas they are even more consonant than they sound. All states legitimize secrecy in the interests of national security. And secretive nations are often the most insecure. Almost by definition, despotic regimes conceal terrible deeds to reinforce the wills of tyrants. But no nation confesses its sins as thoroughly as it professes its virtues. The United States of America, that glimmering beacon of freedom and democracy, almost certainly keeps more secrets than any other nation.

According to a count by the historian Peter Galison, the United States has compiled enough secret documents in the past 80 years to overflow every library in the world. The government, as far as we know, regularly issues some three times as many classified records as unclassified ones, and it now spends roughly $20 billion a year just on classifications. These secrets have not always made its citizens more secure. Even a peek at a classified document from the advent of the modern security regime, during the early Cold War, is enough to reveal the folly.


In 1951, a baker’s dozen of congressmen and their executive staff were bothered by the lack of security inside the US Capitol Building. At most, one member of the overstretched Capitol Police Force patrolled these congressmen’s chambers, in which a vault with an electronic alarm protected some of the nation’s greatest secrets. During nights and weekends, a single officer lolled in the hallway outside to ensure that no one entered the exclusion area within.

One morning in early May, the congressmen and their aides met and debated the poor security. One guard could be “slugged” or “overpowered rather easily,” they complained. And the officers on duty had no special training or loyalty to the congressmen. After all, the director of the committee’s staff confided, the Capitol Police Force was largely part-time and “perhaps not the most superlative in the United States.”

So, the director advised the congressmen to employ their own detail of 10 men, trained by one of the intelligence services, with a minimum of two guards on duty at any one time. The director argued, “[Y]ou reduce the chance of an inside job, based on the connivance of a policeman, when you have two guards on hand rather than one.” And in the unlikely event that the Capitol were ever stormed by “say groups of three, four, or five,” two guards would have a better chance of calling for backup than one. The guards might even be outfitted with “a walkie-talkie business like they can wear in their neckties,” one senator said. The congressmen agreed but balked at the cost of so much security: $50,000.

The congressmen and their aides were debating the Capitol security only because they were plotting to acquire one of the tippy top secrets in the United States. They intended to request the numbers, kinds, and locations of the country’s nuclear weapons from the Department of Defense. As members of the Joint Committee on Atomic Energy, which regulated nuclear arms and energy, they had previously asked for these secrets, as was their right by law. How else could they evaluate the weapons program and provide for the national security?

“They can’t duck behind secrecy,” one senator scoffed. Yet they did. General George Marshall, the Secretary of Defense, had rebuffed the earlier congressional inquiry because he, too, fretted about the nation’s security. Revealing the secrets of the stockpile, even to regulators with top-secret clearances, would endanger the nation, which was then at war.

But in the top-secret transcript of their meeting, partially declassified 40 years later, the congressmen were better informed than they ever admitted to their rivals. The United States, they knew, had begun transporting atomic bombs to “territory completely within our control,” even to foreign countries such as Japan, to better reach its enemies. The congressmen were worried that the military might inadvertently reveal this secret, because its officers could not regulate themselves.

Three weeks earlier, at an airport somewhere along the seaboard, so many armed soldiers had greeted an airplane that reporters thought President Dwight Eisenhower was aboard. When neither the president nor anyone else of rank descended, reporters guessed the true cargo. “Anybody whose head was half hot would know that something of the greatest importance was there,” one senator said. Excess security might expose the other atomic bombs stashed around the world.

The congressmen decided to make a play for the nation’s top secrets in order to guard them more tightly inside their vault. But the Department of Defense had already refused to disclose these secrets to ensure security. Of course, these men were not merely wrestling over national defense. Secrets conferred power and control, the currencies common to politicians and soldiers. Both were covens of mythomaniacs.

The congressmen resolved to press the Secretary of Defense for information, but they did not make a decision to increase security before they adjourned. They went off the record, even during their top-secret meeting, before moving on to other items. They debated whether they could ever trust experts and scientists, especially the nation’s top physicists, who agreed that the Soviet Union had detonated its first atomic bomb. The congressmen then argued about emoluments for public servants.


I rarely consult national archives for their reverberations today. But I stumbled onto this document while I was reading Restricted Data: The History of Nuclear Secrecy in the United States, by the science historian Alex Wellerstein, because I had wondered whether there were any half-secrets left to find. Wellerstein had seemingly found them all and recounted them in his masterful book. While I was reading the document I found, however, I remembered the last time I had sought an historical source for a better understanding of our day.

A week or so after the presidential election in 2016, still in the throes of my hangover, I had visited Wellerstein’s blog, also called Restricted Data. I occasionally lurk on the site because he has a facility for hypotheticals and because so many keen historians are among his commenters. Sure enough, 10 days after the election, Wellerstein addressed my concern for Armageddon. He retold, clearly and horrifically, the history of why there are so few checks on a president’s power to deploy nuclear weapons. When I finished reading, I returned to my drinking.

Many years ago, I audited a graduate course in the history of science, and the deft professor offered students this reasonable advice for success: display mastery over a single event, a single person, a single topic. Never, ever, attempt to tell the history of an entire era or subject. Wellerstein never heeded such advice. He finished his stellar graduate thesis more than a decade ago, and in the introduction to his book, which evolved from that thesis, he readily admits the difficulties of his research. Only a certain kind of person, both foolish and resolute, would choose to study a subject so extensive, yet so restrictive, as the secrets of nuclear weapons. Too many impediments are intrinsic to the task. Secrets, after all, are secrets. Their censorship and classification are blockades to any historian.

All archives have gaps and omissions, Wellerstein explains, but even so: “What makes secrecy feel different is its intentionality: the information I may want is actually knowable and may even be known, but just not by me, at least right now.” He could have submitted to a security clearance to be in the know, but he elected not to. “This no doubt leaves many additional gaps in the story, but it also allows me to share what I have found with impunity.”

Share he does. Like drunks and essayists, government classifiers sometimes disclose too much, owing to overwork and the technical nature of their enterprise. Wellerstein is also an archive rat, who has seemingly built a nest from every document available to him. The gaps he has to leap over are merely extra degrees of difficulty that add to his impressive score.


The history of secrecy and security in the United States is in part the story of science and technology. In Restricted Data, Wellerstein has drafted one of the finest blueprints of our national security apparatus by focusing on nuclear weapons, its deepest cogs and wheels. He reveals the wiles, machinations, and ruses of physicists who first kept the secrets of the nucleus. He uncovers the prevarications, leaks, and conspiracies of the officers and bureaucrats who held those physicists to account. He has found a peephole into a stadium where the most important games are played.

For as long as humans have lived, humans have lied. In the United States, the government is a reflection of its people, yet equivocations were not formally ingrained into American law until 1911. Secrets always multiply during times of war, and during World War I they duly escalated, owing to the new technologies employed by the US military and fears of espionage. But the classifications of secret documents did not really get going until World War II. And as Wellerstein argues, it was the possibility of even greater technologies of murder, the discovery of nuclear fission and the development of atomic weapons, that helped institutionalize secrecy afterward.

Physicists, as Wellerstein notes, have always been a shifty bunch. They are members of a guild that transmits arcane knowledge through a mathematical code known only to initiates. Even among themselves, physicists can be cagey about their work. They toil for years and reveal only their triumphs to garner the recognition of peers. Scientific societies have long had vaults in which a physicist could store a time-stamped manuscript, to ensure credit for ideas should anyone else make a claim.

Owing to their wiles and craft, physicists are also gossips, blabbing about each other’s work and squabbling over ideas. But they communicate their positive results publicly in the end, so they can commend themselves on their openness. In fact, as Wellerstein explains, quoting the scientist and historian J. D. Bernal, “the growth of modern science coincided with a definite rejection of the idea of secrecy.” Physicists came to believe that science could not progress without candidness.

If openness was the key to scientific progress, however, then secrecy was the key to halting it. In 1939, after the discovery of nuclear fission in Nazi Germany, Leo Szilard insisted that all subsequent research on the process should be kept a secret, because a chain reaction inside a critical mass of uranium would lead to an explosion. As Wellerstein notes: “These tensions, between the ideals of science and secrecy on the one hand, and of desires for openness and security on the other, are what make the history of nuclear secrecy in the United States unpredictable, surprising, and, at times, bizarre.”

Szilard applied for a secret patent and tried to persuade his colleagues to censor their own research. But they were either too fervent about openness or too desirous of credit to conceal their work. So, Szilard then asked leading journals to regulate what information appeared. When the embargo failed, physicists abdicated their responsibility to institutions more adept at secrecy: the US military and government. The physicists, Wellerstein notes, “would quickly become accustomed to working within a secrecy regime.” They were already so furtive that they hardly had to adapt.

Wellerstein asks brilliant questions that reach to the heart of what secrecy and science and security mean:

How could a fact of nature be rendered effectively into a state secret, if any scientist, in any laboratory, in any country, could replicate and rediscover it? […] If secrecy were made the norm, would science thrive, or even survive? Which would serve the nation’s security more, keeping things secret, or racing forward as fast and as openly as possible?

Neither the destructive potential of uranium nor the manner of releasing its potential could ever remain state secrets. Reporters had already revealed that physics too many times. But the industry and engineering that a fission bomb required, especially the scale and seriousness of the Manhattan Project, could be hushed.

Wellerstein takes the reader down the long path to understand what nuclear secrecy meant, guiding the reader through the subject’s many tangles. Even on the well-worn track of the Manhattan Project, he points out new vistas unseen by Richard Rhodes, in his voluminous histories of atomic weapons.

The atomic bomb was the most fatal surprise of any war, at least since the Greeks bestowed an oversized horse on the people of Troy. But roughly half a million men and women worked for the Manhattan Engineer District during the war. And there was so much turnover that only 125,000 people were ever employed at any one time. How was the immense project, encompassing whole cities, ever kept a secret?

The water tower in Hanford, Washington, the site of plutonium production, constantly reminded workers “Silence means Security.” And in Oak Ridge, Tennessee, the so-called Secret City built for uranium enrichment, employees had recreational leagues and pep rallies to keep them happy and quiet. But General Leslie Groves, who oversaw the entire project, also employed security officers to surveil those in the know and those on the outside. He relied on every subterfuge available, as Wellerstein notes: codes, misinformation, background checks, clearances, misdirection, wire taps, isolation, guards, fences, censorship of the press, “need to know” policies, and oaths. Physicists abided by most of these; at Los Alamos they even swore an oath on their “scientific reputations,” rather than their lives, to keep their mouths shut.

Yet, Wellerstein can only guess why President Roosevelt ordered so much secrecy from the project’s inauspicious start. Harry Truman, his last vice president, who had investigated rumors of the secret project as a member of Congress, did not even learn about the atomic bomb until 12 days after becoming president (Joseph Stalin and Soviet scientists already knew more than he did, owing to espionage). One executive prerogative, almost certainly, was money. Congress would have never earmarked funds for an untested weapon based in nuclear physics, especially one that ultimately cost $2 billion. Secret funding helped ensure that no one would be accountable, yet, for the result.

Secrecy also had one by-product that the military appreciated — terror. Groves and the military had always intended to reveal the greatest secret about their endeavor, the atomic bomb itself, but they had wanted to garner the most bang for their pluck. If the weapon were a surprise, it could frighten the enemy into submission.

When President Truman announced the destruction of Hiroshima, many hours after the detonation of Little Boy, he disclosed the locations of the secret installations that had made this terrible weapon. He even discussed the scale of the industry and labor involved. Even so, he insisted, there were secrets to maintain. The United States would not, he assured listeners, “divulge the technical processes of production or all the military applications, pending further examination of possible methods of protecting us and the rest of the world from the danger of sudden destruction.”

Days later, however, General Groves authorized the publication of a report about these processes, written by Henry DeWolf Smyth. The Smyth Report, which was released so quickly that printers forgot to stamp its no-longer-secret title on the cover, would largely credit physicists for the project’s success. The physics of the atomic bomb had become so insignificant to its remaining secrets that Smyth could tout their contributions.


General Groves was so proud of his secrecy regime during the war that military intelligence reprimanded him for boasting of it afterward, fearing he might draw further attention to the nation’s guile. Groves did not yet know the extent of espionage during the war, so he believed that his strict regime was critical to the project’s success and should remain in place. Secrecy about fission and the bomb were thus the first links in the chain reaction that exploded into the modern security state.

After the war, J. Robert Oppenheimer and other notable physicists relished their proximity to power. The perks included access to secrets inside the safes, police escorts, and boozy trips to Washington. But those who fretted the long term, including Oppenheimer, also believed that the only real security was an end to secrecy. They imagined a diplomacy based on the ideals of science, a free and open exchange of ideas in the interest of goodwill. The secrets of the bomb, they knew, were largely feats of engineering. Other nations need only acquire uranium and emulate the Americans in devoting their resources to a project. The US military and most politicians thought scientists naïve. They favored total secrecy. And Groves had already begun hoarding uranium to ensure an American monopoly.

Wellerstein writes: “The postwar system attempted to have everything both ways. Science needed to be open, but the bomb needed to be contained.” Physicists organized and even lobbied against the first legislation that would have retained military control over nuclear energy. They advocated, instead, for civilian control and what would become the Atomic Energy Commission.

Although near to power, physicists did not know how to wield it. Where the archive ends, Wellerstein engages in some credible speculation as to how they were outmaneuvered. While Congress was debating how to regulate the nucleus, General Groves likely revealed that spies had tried to infiltrate the Manhattan Project. He then reinforced the importance of secrets and restrictions, although the possible espionage had happened under his watch. The Atomic Energy Act that Congress passed would thus resemble the earlier legislation that physicists had fought. It even defined a new class of secrets — restricted data — and severe punishments for revealing them. All information about nuclear weapons, as historians have noted, was henceforth “born secret.”

Stanley Kubrick understood that nuclear weapons often revealed the United States and its security at their most absurd. Wellerstein prefers to speak of “irony,” a word that, along with its adjectival and adverbial cousins, he uses more than 25 times. He relates one astonishing irony after another. How the physicist John Wheeler lost classified documents on a train while helping a vengeful cabal declare Oppenheimer a threat to national security. How President Truman approved the development of a hydrogen bomb under an order of absolute secrecy, after a senator revealed the possibility of such a weapon on a television talk show about the subject of secrecy. How even testing a hydrogen bomb could disclose the secrets of its manufacture, in the fallout that gusted around the world. How lawyers for the United States, in the espionage trial against Julius and Ethel Rosenberg, risked calling witnesses who might reveal greater secrets, and did in fact allow one secret of implosion bombs to be an exhibit. How the Atomic Energy Commission was an oppressive bureaucracy of secrecy even as it became overzealous about the free market of atomic energy.

Perhaps the greatest irony of the Cold War, however, was that the United States, to preserve its freedoms at home, developed a security regime to rival that of the Soviet Union. Privacy was a right afforded to individuals, but state secrecy trumped that — and every other — right. All persons who worked at national laboratories, whether involved in secret research or not, underwent FBI background checks into their character and loyalty. So, too, did students who received federal grants. The FBI investigated more than 140,000 people in less than two years, roughly eight every hour.

Secrecy became a weapon to destroy people’s lives. FBI agents, who had no training in science, hounded leftist physicists who carried nuclear physics, and possibly secrets, in their heads. Many lost their jobs and clearances, a few their citizenships. The chief of the Organization and Personnel Division at Oak Ridge was fired because of “homosexual activity.” The Rosenbergs (and others, hints Wellerstein) lost their lives. Furthermore, secrecy could veil the state’s crimes. The Atomic Energy Commission elided the harmful effects of nuclear waste and radiation, especially to Pacific Islanders and native populations in the American Southwest. The Commission also suppressed documents that described how physicians injected terminally ill patients with plutonium to study its physiological effects.


Wellerstein adeptly narrates almost 70 years of secrecy in just 400 pages, although nearly three quarters of Restricted Data covers just 25 years, from the discovery of nuclear fission in the late ’30s to the security trial of J. Robert Oppenheimer in 1954. Even in that span, the nature of secrecy changed. Policy-makers slowly appreciated, as Wellerstein notes, that “information is hard to regulate, control, and verify, but physical and material things are not.” From the plaything of science, the atom became a problem of technology, policy, and international relations. Eventually, however, tests of atomic bombs became so commonplace that they were kitsch, another attraction in Las Vegas when not playing the slots.

After Wellerstein quits the 1950s, his chapters become episodic. He leaps from the development of centrifuges by former Nazi scientists in the Soviet Union to the entrepreneurs who tried to enrich themselves by igniting nuclear fusion with lasers. I have two theories for the sudden lumps in his narrative: there are not as many unclassified documents from later eras, and nuclear secrets became more diffuse and unwieldy, owing to free enterprise.

Although Wellerstein does focus his microscope on the smaller organs of secrecy, he never loses sight of the greater beast. He depicts the many college students who showed the public how to design atomic bombs using unclassified documents, and how the Atomic Energy Commission tried to silence their disclosures. He details the stories of journalists who went even further. During the mid-1970s, a dogged reporter named Howard Morland learned so much about hydrogen bombs, from encyclopedia articles and conversations with physicists, that he wrote an article to disclose one of the lone technical secrets that remained. The Atomic Energy Commission took him to court, in a test of free speech and the press, until Morland discovered that the Commission had already declassified the secret by mistake.

By the late 1970s, after the revelations of so many atrocities perpetrated by the United States “in the name of national security,” activists could argue: “Secrecy was a rot that had infested the US government and was at the core of all its problems.” So, they pried open the nation’s greatest remaining secrets.

During the 1980s, William Arkin, a former army intelligence officer, detailed the American stockpile of nuclear weapons, including the counts and locations, which he gleaned from public documents and by snooping around. But it took a Black woman, Hazel O’Leary, the Secretary of Energy under Bill Clinton, to launch the Openness Initiative, which finally revealed most of the abuses of the nuclear program. Within a decade, the Information Security Oversight Office had unclassified more than 1.1 billion pages of documents. Although the Office still reviews about 100 million pages every year, the production of secret documents exceeds that rate. In 2020, a report suggested that the United States would have to turn to artificial intelligence if it ever hoped to catch up.

Restricted Data only fizzles once, briefly, in a petite final chapter that describes some 20 years of history, from the end of the Cold War to the presidency of Barack Obama. Wellerstein rushes through the nuclear weapons programs of North Korea, Pakistan, Israel, Iraq, and Iran, often in single paragraphs, and he almost completely ignores the most recent US presidency. I would be a harsh critic to fault him for not being completist. In the few days when I was reading his book, I noticed just how often nuclear energy and arms still comes up: Russia threatened the United States with proliferation again, the 10th anniversary of the disaster at the Daiichi nuclear plant in Fukushima passed, and experts warned that nuclear weapons remained the greatest threat to humanity.

More than 70 years ago, physicists gave humans the means, if not the motivation, to slaughter all life on earth. That we have not done so yet is hardly the triumph of secrecy. Wellerstein even concludes that “the value of ‘secrets’ appears overblown.” Most security experts have reached the same conclusion. For the latest threats to humanity, cyberweapons, “researchers do not believe secrecy is an effective countermeasure, and tend to prefer radical openness, because these informational ‘weapons’ can in fact be countered by other informational countermeasures.”

In 2010, in the interests of openness and national security, the Department of Defense finally released the information that it had denied to congressmen 60 years earlier. The number of warheads in the nuclear stockpile was a little more than 5,000. Few noticed and fewer cared. Nowadays, congressmen are more preoccupied with mobs storming the Capitol for imaginary secrets and absurd conspiracies that have nothing to do with nuclear weapons.


Joshua Roebke is a writer and historian of science. His first book, a social and cultural history of particle physics, is forthcoming.

LARB Contributor

Joshua Roebke is an author and historian of science with a background in Spanish literature and theoretical physics. He is finishing his first book, a social and cultural history of particle physics titled The Invisible World, which won a Whiting Foundation Creative Nonfiction Grant. He was an editor at an award-winning science and culture magazine, and his writing has appeared in Joyland, Scientific American, Wired, Quanta, Salon, Kenyon Review, Undark, and elsewhere. One of his articles appeared in The Best American Science and Nature Writing and another was nominated for a Pushcart Prize. He teaches literature and writing in the College of Natural Sciences at the University of Texas at Austin.


LARB Staff Recommendations

Did you know LARB is a reader-supported nonprofit?

LARB publishes daily without a paywall as part of our mission to make rigorous, incisive, and engaging writing on every aspect of literature, culture, and the arts freely accessible to the public. Help us continue this work with your tax-deductible donation today!