The Pentagon's Brain
“Nonlethal weapons are generally intended to prevent an individual from engaging in undesirable acts,” wrote E. E. Westbrook and L. W. Williams, the authors of the report. “Apart from the moral arguments in the present and future use of nonlethal weapons, public officials find it prudent to examine nonlethal force using a framework that it was keeping ‘innocent bystanders’ from being hurt.” At the overseas CDTCs, ARPA chemists examined a variety of incapacitating agents for future use against protesters, including dangerous chemical agents with a wide range of effects, from vomiting to skin injury to temporary paralysis.
Possible irritants for use against demonstrators included “CN (tear gas)… CS (riot control agent)… CX (blister agent),” also called phosgene oxide—a potent chemical weapon that causes temporary blindness, lesions on the lungs, and rapid local tissue death. CS was seen as a viable option: more than 15 million pounds of CS had already been used in Vietnam to flush Vietcong out of underground tunnels on the Ho Chi Minh Trail. CX was also recommended for crowd control. It “produced a corrosive injury to the skin, including tissue injury,” but because the worst damage was inside the lungs, the harm would be disguised. Anticholinergics were considered, chemicals that cause physical collapse. “Probably the most promising of the anticholinergics (agents which block passage of impulses through parasympathetic nerves),” wrote the chemists, were compounds that produced “rapid heart rate, incoordination, blurred vision, delirium, vomiting, and in cases of higher doses, coma.” Emetic agents, chemicals that induce vomiting, were also recommended.
A second program involved delivery systems. Mechanisms for delivery included liquid stream projectors, able to shoot a twelve-inch-diameter stream of liquid across a distance of up to forty feet, as well as grenades thrown by hand or discharged from a small rocket. A more powerful option was the E8-CS man-portable tactical launcher and cartridge, which could be fired electrically or manually into rioting crowds at a distance of up to 750 feet. “It is nonlethal in the impact area, but its high muzzle velocity creates a lethal hazard at the muzzle during firing,” the scientists wrote.
Poison darts were discussed as a possible “means for injecting an enemy [i.e., a protester] with an incapacitating agent.” Also recommended were tranquilizing darts, historically effective in subduing wild or frightened animals. The problem, the ARPA chemists cautioned, was that “using these kinds of darts was not entirely safe as accurate dosage was based on the weight of the animal.” One advantage was that the “use of a dart allows selection of an individual target, perhaps the leader of a group or a particularly destructive person, without injuring others around him.” Further, the darts “possess a psychological advantage not shared by many other systems,” noted the scientists. “The victim may wonder what he has been hit with and whether or not it is essential that he find an antidote.” This benefit needed to be weighed against another danger, however, which was that if someone was hit in the head or neck, it could be fatal. “Darts are not regarded by many as an ‘acceptable’ weapon,” the scientists wrote. Following the dart discussion was a long treatise on whether or not the use of the electric cattle prod against human protesters would be defensible.
The 130-page report offered hundreds of additional development ideas about how to incapacitate demonstrators without killing them, programs that were currently being researched for battlefield use but had not yet been deployed in Vietnam. “Photic driving” was a phenomenon whereby the application of stroboscopic light within a certain frequency range could cause a person’s brain waves “to become entrained to the same frequency as the flashing light.” But early studies showed that this kind of flickering light was effective in only about 30 percent of the population. Laser radiation was suggested as a potential way of temporarily blinding people, also called flash blindness. One drawback, the ARPA scientists noted, was that “the laser must be aimed directly at the eye,” which “diminishes its practicality in a confrontation situation.” Microwaves could potentially be used to incapacitate individuals by burning their skin, but the science had not yet been adequately advanced. “Surface skin burns using microwaves would not form soon enough to create tactical advantage,” the scientists wrote. Also, trying to burn someone with a microwave beam would be “ineffective against a person who is wearing heavy clothing or who is behind an object,” the scientists wrote.
Another series of tests researched “the use of loud noises to scare people or to interfere with communications.” But the ARPA scientists cautioned that sound would have to be “so offensive and repugnant that hearers leave the scene,” meaning a volume so high that it presented the danger of permanent hearing loss. “Most subjects experience pain at about 140 db [decibels], and at about 160 db, the eardrum is torn.”
Tagging was an option, to help police make arrests after a demonstration. “The marking of people for later apprehension is another technique which has been tried in some situations,” the scientists wrote, suggesting specific materials including “invisible markings which were sensitive to ultraviolet light” and “odor identifying markings, sensed by dogs or gas chromatographs.”
Crowd control had long been an engineering challenge at the Pentagon. To be effective, nonlethal weapons need to deliver enough power to produce a dispersal effect but not enough power to cause serious injury or harm. Most historical accounts of the use of nonlethal weapons in the United States cite the Omnibus Crime Control and Safe Streets Act of 1968 as a turning point. The act established the Law Enforcement Assistance Administration (LEAA), a federal agency within the U.S. Department of Justice designed to assist state police forces across the nation in upgrading their riot control hardware and officer-training programs. The act also provided $12 billion in funding over a period of ten years. Police forces across America began upgrading their military-style equipment to include riot control systems, helicopters, grenade launchers, and machine guns. The LEAA famously gave birth to the special weapons and tactics concept, or SWAT, with the first units created in Los Angeles in the late 1960s. “These units,” says an LAPD historian, “provided security for police facilities during civil unrest.” But what has not been established before this book is that much of this equipment was researched and developed by ARPA in the jungles of Vietnam and Thailand during the Vietnam War.
In America, antiwar protests raged on. Not even computers could escape the hostility between the Pentagon and the antiwar establishment. In early 1970, a Defense Department computer at the University of Illinois at Urbana-Champaign, called the ILLIAC IV, came under fire. ILLIAC IV was the fastest computer on earth at the time. The scientist in charge of the project for ARPA was Professor Daniel L. Slotnick, a mathematician and computer architect. A former student of John von Neumann, Slotnick had worked with von Neumann on MANIAC, at the Institute for Advanced Study at Princeton, starting in 1952. It was there that Slotnick developed his first thoughts about centrally controlled parallel computers. A pioneer in his field, Slotnick was one of the first to develop the concept of parallel computing, a form of computation in which multiple calculations are carried out simultaneously by separate computers and solved concurrently. Slotnick co-authored the first paper on the subject, in 1958. His goal with ILLIAC IV was to build a machine that could perform a billion instructions per second. Although it used the same architecture conceived by John von Neumann, ILLIAC IV was a far cry from MANIAC in terms of computing power.
ILLIAC IV was fifty feet long, ten feet tall, and eight feet wide. The machine’s power supply units were so massive they had to be moved with a specially designed forklift. The supercomputer was made up of a group of sixty-four processor elements, with a potential for up to 256—a groundbreaking number of processing units at the time. The machine was designed to cut down exponentially on the time it took to complete basic computational science and engineering tasks. Approximately two-thirds of the computer’s time was designated for work on Department of Defense weapons programs, including “computational requirements for ballis
tic missile defense.” Specifically, the calculations sought to differentiate a missile from the background noise, the problem that had been plaguing the Jason scientists since they first began studying the topic in 1960. ILLIAC IV was also used for climate modeling, and for weather modification schemes, as part of a still-classified ARPA program called Nile Blue. Not until July 1972 would the U.S. government renounce the use of climate modification techniques for hostile purposes. In May 1977 an international treaty, the Convention on the Prohibition of Military or Any Other Hostile Use of Environmental Modification Techniques, would be signed, in Geneva, by forty-eight nations. Until then, weather modification schemes were pursued.
Slotnick and his team called the ILLIAC IV “the ultimate number cruncher.” ARPA officials believed that if they had two of these computers, their capability would cover “all the computational requirements on planet earth.” The building of ILLIAC IV, most of which was done by graduate students, was the largest and most lucrative Defense Department contract in the history of the University of Illinois. By late 1969, the university had received more than $24 million in funds, roughly $155 million in 2015. Plans for a fancy new facility to house the machine were in place, with groundbreaking ceremonies to begin sometime during the following year. The specifics of the arrangement between Slotnick and ARPA were classified, but it was not a secret that a supercomputer was being built at the university.
What it would be used for was obscure until January 5, 1970, when the Illinois Board of Higher Education met for a budget review and a student reporter managed to attend. The following day, on January 6, 1970, a headline in the Daily Illini declared, “Department of Defense to employ UI [University of Illinois] computer for nuclear weaponry.”
The revelation that the university was working with the Defense Department on nuclear weapons work had an explosive effect on an already charged student body. “The University has proven that it is not a neutral institution,” declared the antiwar group Radical Union, “but is actively supporting the efforts of the military-industrial complex.” One article after the next alleged malevolent intentions on the part of Professor Slotnick and the dean of the Graduate College, Daniel Alpert, in having tried to conceal from the student body the true nature of the computer. “The horrors ILLIAC IV may loose on the world through [the] hands of military leaders of this nation” could not be underestimated, the Daily Illini editorialized. “We fear the military… will use the computer to develop more ways to kill people and spend the people’s money.” In another article, a group of concerned students wrote, “Considering the evil demonstrated by our military in recent years, we would rather have seen the University resistant to the evil… than complicit with it.”
Professor Slotnick tried to justify the Pentagon funding by pointing out that other institutions were unwilling to fund such an important but far-sighted program as building this supercomputer. “If I could have gotten $30 million from the Red Cross, I would not have messed with the DoD,” Slotnick said. ARPA took offense, calling Slotnick a “volatile visionary.” The board tried to throw a blanket over the fire by declaring the “more important” parts of the computer “non-military.” Despite attempts to humanize the machine, the debate only grew. A teach-in was organized against ILLIAC IV. Students wanted the machine gone.
On February 23, 1970, the protests took a violent turn when unknown persons firebombed the campus armory, causing $2,000 worth of damage. Then on March 2, five hundred protesters disrupted a job-recruiting session with General Electric, the defense contractor that helped build ILLIAC IV. Windows were broken and three people were injured. Officers who tried to arrest people were pummeled with mud balls. The crowd grew to as many as three thousand. When antiwar demonstrators broke windows in the chancellor’s office, state police wearing full riot gear appeared on the scene. Not until late that night was peace restored. Twenty-one students were arrested, eight seriously injured. On March 9, the university’s faculty senate took a vote to oppose ILLIAC IV; it failed. Two days later, the Air Force recruiting station in Urbana was firebombed, the sixth local arson attack of the month.
The spring of 1970 was a tempestuous time on college campuses across America. On April 30, 1970, President Nixon went on national television to announce the U.S. invasion of Cambodia, yet another expansion of the Vietnam War. Nixon’s disclosure that 150,000 more soldiers would now be drafted sparked major protests across the nation. Four days later, on May 4, four students at Kent State University in Ohio were shot dead by the National Guard.
The following day, the ILLIAC IV protests at the University of Illinois ratcheted up even further when two thousand demonstrators stoned police vehicles parked on campus. On the morning of May 6 the National Guard moved in, and on May 7, ten thousand students and faculty held a peace rally. When the university refused to fly flags at half-mast for the victims of the Kent State shootings, students pulled down the American flag that had been flying on the university fire station flagpole and set it on fire. On May 9, demonstrators staged a sit-in in front of the building that housed the ILLIAC IV. Protests and arrests continued until May 12.
In June, university officials told ARPA that they could no longer guarantee the safety of its supercomputer. ARPA began looking for a new facility to house the ILLIAC IV and in 1971 entered into a new contract with a federal research facility in California. Each side—the protesters and the government—believed strongly in the legitimacy of its position. Students at the University of Illinois and elsewhere across the nation continued to protest against war; the Department of Defense continued its weapons research and its war in Vietnam.
The supercomputer was packed up and taken to California. By the spring of 1972, ILLIAC IV was up and running at NASA’s Institute for Advanced Computation at the Ames Research Center. This was adjacent to the U.S. Navy’s west coast facility where highly classified antisubmarine warfare work was taking place. ILLIAC IV began making calculations for the Navy’s Project Seaguard, a classified program to track submarines using acoustics, another ARPA program, with research taking place at ARPA’s classified Acoustic Research Center, deep underwater in a lake in northern Idaho.
The submarine research facility was one of ARPA’s best-kept secrets, an underwater test site located at the south end of a small resort community on Lake Pend Oreille in Bayview, Idaho. The forty-three-mile-long lake is 1,150 feet deep in places, making it the perfect locale to conduct secret submarine research. Acoustic sensors placed on the floor of the lake recorded and processed data which were then fed into ILLIAC IV, allowing for major Cold War advances in antisubmarine warfare.
The ILLIAC IV controversy coincided with a major turning point in the history of the Advanced Research Projects Agency. Public opposition to the Vietnam War, coupled with rising inflation, put an unwelcome spotlight on ARPA when Senator Mike Mansfield, an antiwar Democrat from Montana, introduced a bill that barred the Defense Department from using funds “to carry out any research project or study unless the project or study had a direct relationship to [a] specific military function.” The Mansfield Amendment, introduced in late 1969 as an amendment to the Military Authorization Act, focused “the public’s desire for practical outcomes” against the idea that not only was the Pentagon failing to end the war in Vietnam, but also its spending was out of control. The amendment put military research and development under intense scrutiny and had a direct impact on ARPA. Because most of its work was speculative, looking ten to twenty-five years into the future, directors of the agency would now have to present much more detailed information to Congress before their budgets could be passed.
Then in February 1970 came another devastating blow for ARPA. The secretary of defense authorized a decision that the entire agency was to be removed from its coveted office space inside the Pentagon to a lackluster office building in the Rosslyn district of Arlington, Virginia, two and a half miles away. Desks, chairs, file cabinets, and furniture were all boxed up and moved.
The Pentagon
was the seat of military might, the locus of power. Moving even a short distance away was, as one insider put it, “the epitome of the Agency’s downgrading.” The underlying message being sent to staff was that the Advanced Research Projects Agency might just fold. Even the ARPA director at the time, the electrical engineer and telecommunications expert Eberhardt Rechtin, appeared to have lost confidence in the agency he was in charge of. Rechtin confided to a colleague, “It wouldn’t surprise me that all of a sudden [a secretary of defense] would decide to kill ARPA.” Since its inception in 1958, ARPA had been a place where there was always more money than ideas. Suddenly, “the dollar situation was so bad, [the agency] had far more ideas than money,” Rechtin said. Without money, there was less power, and without power, there was greater tension.
To many on the ARPA staff, it seemed as if Rechtin did not particularly care whether the agency survived. “The staff just didn’t know what was going to happen next,” one program manager told a government historian in 1974. “They didn’t know who was boss. They didn’t know who to follow. They didn’t know whether anyone cared.” The staffer continued: “At least if you kill something[,] you know. You line it up against the wall, you take aim, you spend five minutes at the job and you kill it right. But to let it wither away by not even allowing it to have a Director [who cared] is almost [worse]. The feeling was: he [Rechtin] doesn’t care anymore… he is selling us down the river… we’ve become the pawn, and we are moving away from the center.” An “apocalyptic feeling” overwhelmed the ARPA staff. “We had terrible feelings that this [was] the end,” said another unidentified staffer.