Teller could make that argument because there was, in 1961, remarkably little agreement on whether it would be possible to detect a covert Soviet test, particularly one conducted underground. Both sides had seismic sensors that could detect the rumblings caused by such a test, but it was not clear that scientists could reliably distinguish between natural events, like earthquakes, and the explosion of a nuclear device underground. In other words, either side could simply deny there was a nuclear test and claim what had taken place was an earth tremor. This issue would become key to any test ban negotiations between the Soviet Union and the United States. The debate had become such a highly charged issue for treaty negotiations that the common joke was that Kennedy and Khrushchev had both become amateur seismologists.
Not everyone was so convinced of Teller’s assertions. ARPA was assigned nuclear test detection under the code name Vela at the end of 1959 as a counterweight to the CIA’s and the air force’s secret test detection network. ARPA got the work, quite simply, because President Eisenhower did not trust his spooks and wanted an assessment that was independent of the CIA and its assets. The work at ARPA initially foundered, but the election of Kennedy, who was interested in arms control, brought renewed focus and funding to the Vela test detection program. By 1961, Vela had three parts: Vela Uniform, to detect underground nuclear tests; Vela Sierra, to detect nuclear explosions in the atmosphere; and Vela Hotel, which would launch satellites with sensors to detect nuclear tests from space.
In 1961, ARPA began pouring money into academic seismology research in the United States and abroad. The academic discipline of seismology, at the time, was a backwater. Robert Frosch, who was recruited to ARPA to run Vela, recalled going with the director, Robert Sproull, to visit what was supposed to be a state-of-the-art seismic vault, one of the underground bunker-like structures that were used to measure tremors. The two men came out of the vault in shock, feeling as if they had just emerged from a time capsule. The seismologists there were using pen recorders and primitive galvanometers, an analog instrument used to measure electrical current. Vela began to change that with an influx of funding for seismology that was almost unimaginable in scale for most areas of science. The military’s need to distinguish earthquakes from nuclear tests brought seismology “kicking and screaming” into the twentieth century, according to Frosch. At one point, he said, he funded almost “every seismologist in the world, except for two Jesuits at Fordham University” who refused to take money from the Pentagon.
Frosch’s ambitious idea for advancing both seismology and nuclear test detection was to build a novel system that would identify the vast majority of Soviet earthquakes, resolving once and for all the debate over distinguishing earth tremors from nuclear tests. Frosch’s project, called the Large Aperture Seismic Array, or LASA, was a massive nuclear detection system that comprised two hundred “seismic vaults” buried across a two-hundred-kilometer-diameter area in the eastern half of Montana. For it to work, more than a dozen of these enormous sites would have to be constructed around the world to monitor the Soviet Union. There had been smaller arrays, including one in the United Kingdom, but no one had ever built an array of LASA’s size or scope or knew if it would really improve detection. The air force hated the idea, and seismologists, Frosch said, considered his idea “mildly crazy.” Frosch saw it as a way of putting ARPA’s flexibility to the test. “If you had an idea, you didn’t have to go through two years of getting permission and three years of getting the contract people to make a mess of it,” he said.
ARPA’s leadership gave it their blessing. In the end, it required negotiating with some fourteen different utility cooperatives and dozens of Montana landowners, who were less than thrilled to have the federal government installing nuclear detection devices on their private property. Frosch recalled ARPA getting a complaint from a landowner, who spotted personnel one morning working on a seismic vault. “When I was sitting there eating my breakfast yesterday, I could see someone digging on my land,” the owner reportedly said. “I don’t like to see someone on my land when I’m eating my breakfast.”
What was amazing about LASA, according to Frosch, was the scale of the work, which was completed in just eighteen months, a schedule unimaginable for government projects that typically take years, if not decades. When ARPA needed to have a center where all the seismic data could be collected and analyzed, the agency ended up renting space in downtown Billings, where data from the array was routed to an IBM computer. “It was literally opened up in a storefront in Billings, and in the back room we had computers and we were the center of the array,” Frosch said.
ARPA also began funding the placement of seismograph stations around the world that were operated by scientists. The ARPA-funded seismograph network was designed not to replace the military’s classified system but merely to expand the science of test detection. It was, however, a powerful counterweight to the CIA and the air force, who up to that point had a monopoly on advice to political leaders about what was theoretically possible to monitor in a test ban. Naturally, the CIA and the air force regarded ARPA as “a bunch of incompetents,” and worse, incompetents who were “insisting that their work be placed in the public domain.”
The Worldwide Standardized Seismograph Network, as the ARPA project was called, was massive in scale and scope. Instead of recording results on paper, these new stations would convert results onto 70 mm film, which could be more easily transported and shared among researchers. That was one revelation for seismologists; the second was the extraordinary reach of the seismographic stations, which stretched from Alaska to Tasmania, often in remote or exotic locations. In Trieste, Italy, mountain climbers helped place the seismometers; other locales involved using dogsleds and rickshaws. ARPA also encountered setbacks, like in the South Pole, where instruments ended up frozen inside a block of ice.
According to Lee Huff, who co-authored the early ARPA institutional history, it fell to Godel and his team to secure the agreements to install some of the initial seismic arrays. They began to fan out across the globe to negotiate ARPA’s new system, from Thailand to Iran. “Godel had already a lot of experience putting in networks of listening devices of one kind or another around the Soviet Union. He did some of that for the NSA, literally, going himself out into the countries and getting it done,” Huff said. “He was an old hand at that.”
—
For the most part, getting a country, be it India or Iran, to agree to the seismic stations was not difficult at the time. ARPA was essentially offering to build—for free—seismological stations that would be operated by the host country, and local scientists only needed to agree to operate them and share the data. The network would eventually involve some 125 stations in more than sixty countries. “You could do things a lot easier than you can today, and they were simply letter agreements,” said Jon Peterson, a scientist for the U.S. Coast and Geodetic Survey, a federal agency that worked with ARPA on the stations.
As ARPA built up a worldwide detection network, the program highlighted a growing tension between secret and open research that was about to become critical for arms control and national policy. The air force and the CIA refused to release data from their network of sensors. “Everybody thought that they were too secretive, they always gave you data that you wanted, but they wouldn’t open up their books,” Ruina said. The bête noire of the nuclear detection world was Carl Romney, a scientist who worked for the Air Force Technical Applications Center, or AFTAC, the agency responsible for nuclear test detection. Romney was widely regarded as the nation’s leading expert in test detection. Many in the field called him brilliant, but he was also widely vilified by critics for his role in blocking every attempt to move forward with a test ban. “Romney never tried to mess up the data, he would only deliberately misinterpret the data,” claimed Jack Evernden, a seismologist who worked for Romney, before later moving to ARPA.
Whether deliberate or not, the problem with secret data, as Ruina pointed out,
was that “nobody could argue with it; they could just question it.” The secret data problem came to a head in 1962, when the United States carried out a test called Aardvark, a part of the first series of tests conducted completely underground. Aardvark, a forty-kiloton nuclear device intended for nuclear artillery, produced reliable seismographic data on a nuclear underground explosion, and Romney suddenly realized he had been wrong about a critical national security issue. He had been arguing that it would be difficult to distinguish small underground nuclear tests from earthquakes, which would make verifying a nuclear test ban treaty difficult, if not impossible. Now, with the Aardvark data, he knew he had been wrong on a key point. During a July 3, 1962, meeting, Romney announced that the new seismic data led him to conclude that distinguishing between tremors and small nuclear tests might not be as difficult as he had previously thought. McNamara was upset and decided the Pentagon had to issue a press release, because if the new data leaked—which often happened—it would look as if the government were “withholding information that would tend to ease the inspection problem in a nuclear test ban.”
Ruina called it an “honest mistake,” but one that would have been avoided if other scientists had been given access to the classified data that Romney jealously guarded. “This is what can happen when you have one person interpreting data, there’s no peer group reviewing it, and there’s nobody duplicating the experiment,” the ARPA director wrote in a three-page letter, blaming the mistake on secrecy. ARPA’s work finally began to sway President Kennedy, according to Glenn Seaborg, who as chairman of the Atomic Energy Commission played a key role in test ban negotiations. Kennedy paid close attention to ARPA’s test detection research, relying on its results in deciding how to go forward with the treaty. “VELA seemed to indicate that the detection capability was better than had been thought by American experts in the period from 1959 to 1961,” Seaborg wrote in his memoir detailing the negotiations.
On October 7, 1963, a month after Ruina left ARPA, President Kennedy signed the Limited Test Ban Treaty, which halted nuclear testing in the atmosphere, in outer space, and underwater. The first launch of ARPA’s Vela Hotel satellite, to detect space-based nuclear tests, took place just days later. It was, according to all accounts, a spectacular success, proving the earlier naysayers wrong. Though the treaty did not prohibit underground testing, the ARPA work was credited with helping reach any agreement with the Soviet Union. “There are three reasons for the partial test ban,” said Robert Sproull, who took over as director after Ruina left. “One that Mr. Kennedy wanted it, one that the Soviet Union wanted it. And one that ARPA made it possible for the Senate to ratify it. And all three of those were required.”
—
By 1963, ARPA’s Vela program had established a counterweight to the intelligence community and, in the process, helped achieve a limited test ban by proving that detection of atmospheric and underground nuclear tests was indeed possible. Vela would also over the next few years have an equally significant impact on the science of seismology. As measurements from the Worldwide Standardized Seismograph Network started coming in and being shared among academics, the seismologist Lynn Sykes was able to use the data from ARPA-funded stations to more accurately track the location of oceanic earthquakes. Whereas seismologists once placed the earthquakes all over the ocean floors, Sykes was now able to prove that in fact the tremors were occurring along mid-oceanic ridges. Plate tectonics, which had previously been a highly controversial theory, could now be substantiated with data from ARPA’s network.
In 1968, Sykes, along with his fellow seismologists Bryan Isacks and Jack Oliver, published a seminal paper that finally paved the way for the acceptance of plate tectonics. “Seismology and the New Global Tectonics” relied heavily on years of results drawn from the ARPA network, including data that showed how waves crossed within and between continental plates. As Sykes put it, Vela “almost instantaneously transformed seismology from a sleepy, poorly supported scientific backwater to a field flooded with new funds, professionals, students and excitement.”
—
Sometimes science and policy meet up, as they did with arms control in the early 1960s; other times they do not, demonstrating that technology alone is not enough to solve problems. After the Limited Test Ban Treaty of 1963, enthusiasm for further negotiations waned. Following Kennedy’s death in 1963, Lyndon Johnson expressed little interest in pushing forward with a comprehensive test ban. It would take almost three decades for the United States to declare a moratorium on underground nuclear testing, in 1992, which followed a unilateral Soviet moratorium the year prior. The Senate never ratified the Comprehensive Nuclear Test Ban Treaty, even after the agreement was adopted in 1996. Nonetheless, Vela is widely—and properly—credited as a technical and political success for its early contributions to a limited test ban and then for helping move along negotiations for a comprehensive ban, even if belatedly. In the process, Vela also “revolutionized seismology,” concluded Lee Huff.
ARPA’s legacy in missile defense, on the other hand, was far more tenuous. The Arecibo Observatory that briefly interested the NSA proved to be an excellent science facility but contributed almost nothing to the military. The megadeath particle beam known as Seesaw did not produce anything of particular value for the Pentagon. And ARPA’s advocacy for exotic directed-energy schemes for missile defense inadvertently planted the seed that would blossom decades later under President Ronald Reagan as a fantastical global shield. ARPA could, however, take credit for killing off some “loony” ideas like BAMBI’s killer spiderweb and dissuading senior leaders from pursuing doomed efforts like Nike Zeus.
Yet ARPA’s work in the nuclear world produced new science and advanced technology, enabling political change. It represented a “triumph” in technology and national security, according to Stephen Lukasik, who was recruited to ARPA in 1966 to work on nuclear test detection and later became the agency’s director. It was a triumph because the science matched the politics, he argued. Ballistic missile defense and test ban treaties were at the top of the White House’s agenda, and ARPA’s scientific progress in these fields marched in lockstep. It proved that verification was scientifically feasible and that missile defense was technologically infeasible. ARPA was successful, according to Lukasik, “not because we were geniuses—we were just ordinary geniuses—but because the country was ready for test bans, the country was ready for nonproliferation treaties, the country was ready for ballistic missile defense treaties, and the country was ready for limitations on theorized nuclear weapons and cruise missiles.”
In other words, ordinary geniuses transformed an entire scientific field and helped open the door to arms control, not just because ARPA scientists had the freedom to do what they wanted, but because these ordinary geniuses were working on problems of national importance. The question was what ARPA could accomplish if it employed an extraordinary genius.
CHAPTER 7
Extraordinary Genius
“We have some big trouble,” President John F. Kennedy told his brother Attorney General Bobby Kennedy early in the morning of October 16, 1962.
A few hours later, the younger Kennedy was staring at pictures of Cuba that had been taken by the U-2 surveillance aircraft. “Those sons of bitches, Russians,” he said, sitting in the White House with a group of officials dedicated to overthrowing Fidel Castro.
The pictures showed the telltale signs of Soviet missile launchers. The CIA had used a massive computer—it took the better part of a room to house it—to calculate the precise measurements and capabilities of the missiles installed. Their dismal conclusion was that the missiles had a range of more than a thousand miles, making them capable of reaching Washington in just thirteen minutes. This revelation touched off a crisis that lasted almost two weeks. As the standoff over Cuba intensified, American military forces reached DEFCON 2, just one alert level before the start of a nuclear war.
As military and civilian commanders clamored for information
on a minute-by-minute basis, computers like the air force’s IBM 473L were being used for the first time in the midst of a conflict to process real-time information on how, for example, to allocate military forces. According to a top secret Pentagon report on military command and control, the Cuban missile crisis demonstrated how operational data, now available on computers, “were recognized by more and more Joint Staff officers,” and “informal requests for outputs increased.” Yet even with the growing availability of computers, sharing that information among military commanders involved a time lag. The idea of having the information travel between connected computers did not yet exist.
After thirteen days of scrambling forces to carry out a potential attack, the Soviet Union agreed to remove its missiles from Cuba. Nuclear war was averted, but the standoff also demonstrated the limits of command and control. With the complexities of modern warfare, how can you effectively control your nuclear forces if you cannot share information in real time? Unbeknownst to most of the military’s senior leadership, a relatively low-level scientist had just arrived at the Pentagon to address that problem. The solution he would come up with became the agency’s most famous project, revolutionizing not just military command and control but modern computing.
—
Joseph Carl Robnett Licklider, who went by the initials J.C.R., or simply Lick to his friends, spent much of his time at the Pentagon hiding. In a building where most bureaucrats measured their importance by proximity to the secretary of defense, Licklider was relieved when ARPA assigned him an office in the D Ring, one of the Pentagon’s windowless inner rings. There, he could work in peace and hopefully avoid William Godel, who always seemed to be sticking his nose in everyone’s business or, worse, would try to get Licklider involved in what was inevitably a cockamamie Vietnam project.
The Imagineers of War Page 12