The Imagineers of War

by Sharon Weinberger

  ARPA’s role in Pandora immediately evoked concerns among the few Pentagon scientists who were cleared to review the program. Bruno Augenstein, a German-born physicist who worked for the Defense Department, sent a top secret memo to Harold Brown and Gene Fubini, two of the Pentagon’s top technology officials, to let them know that ARPA was evaluating proposals looking at the neurological effects of microwaves. In his note, Augenstein alluded to “past unsavory history of experiments of this kind in this country, which has made a number of people rather leery of further experiments in this field,” likely a reference to the CIA’s infamous MKULTRA mind control experiments begun in the 1950s, in which agency officials tested the effects of LSD as a possible mind control agent on humans. Augenstein wrote that there did appear “to be some internal resistance in ARPA to the suggestion that ARPA proceed with these experiments, probably because there is a feeling that at one time it certainly attracted a number of crackpots.”

  If the ARPA program was supposed to avoid the mistakes of prior scandals in human experimentation, then Cesaro was an inauspicious choice to lead Pandora. A propulsion expert, he had no apparent expertise in the biological sciences, but he relished running a top secret project that had high-level attention from the White House and the CIA. He embraced the assignment with an enthusiasm that might have been admirable, had it not been quite so morbid. It soon became clear Cesaro’s primary interest was pushing forward with actual microwave weapons, rather than understanding the underlying biology.

  To see if the Moscow Signal really affected human behavior, ARPA first started by testing microwave radiation on monkeys. Because Pandora was top secret, the primary research had to be run at government laboratories rather than at universities. The air force was assigned to provide electromagnetic equipment needed to generate the microwaves, and the Walter Reed Army Institute of Research was responsible for selecting the monkeys and running the experiments. The initial tests were designed to see how primates performed work-related tasks when exposed to radiation that matched the Moscow Signal, which was beaming every day at the men and women inside the American embassy in Moscow.

  The test protocol involved training the monkeys to press certain levers in response to signals. If the monkeys pressed the lever correctly, they would receive a reward of food, “much as embassy employees might be rewarded with a dry martini at the end of the day,” wrote the columnist Jack Anderson. Researchers would then measure whether the monkeys performed worse when subjected to the Moscow Signal, compared with when there was no radiation. By December 1965, shortly after the lab work had started, Cesaro was already enthusing over the results. The normal process for accepting any new, significant scientific phenomenon would have involved submission of the results for peer review, publication in a respected journal, and eventually replication by an independent group. Pandora, on the other hand, operated in the world of classified science, where results were conveyed not by the researchers conducting the experiments but by the manager in charge, in this case Cesaro. In December 1966, Cesaro reported that the first monkey involved in the tests had demonstrated “two repetitive, complete slowdowns and stoppages” as a result of exposure to the Moscow Signal. “There is no question that penetration of the central nervous system has been achieved, either directly or indirectly into that portion of the brain concerned with the changes in the work functions and the effects observed,” Cesaro wrote.

  The radiation results were so convincing to him that he recommended the Pentagon immediately start to investigate “potential weapon applications.” He initiated a new phase of Pandora intended to move toward human testing, taking the ARPA program dangerously close to the very work that Augenstein, the Pentagon scientist, had warned of. Cesaro also wanted to make Pandora even more secretive than it had been previously. “The extremely sensitive nature of the results obtained to date, and their impact on National Security, has resulted in establishing a special access category for all data results and analysis, under codename Bizarre,” he wrote. Bizarre, as it turns out, was an appropriate name for the project, because at this point the number of monkeys involved in the testing stood at one.

  —

  Initially, the Pandora scientific review committee seemed to go along with Cesaro’s enthusiastic proposal to move directly to human testing. The committee even suggested recruiting human subjects from Fort Detrick, Maryland, home to the army’s biological research program (the conscripts assigned to Fort Detrick have been a continuous source of human subjects for Defense Department research for decades; subjects there have been exposed to everything from yellow fever to hallucinogenic drugs). In minutes from a May 12, 1969, meeting to discuss human testing, the Pandora scientific committee discussed plans to move forward with eight human subjects. The human subjects would be exposed to the Moscow Signal and then given a full battery of medical and psychological tests.

  The committee was aware of the potential for a conflict of interest involved in classified human testing; the idea of informed consent becomes hazy when the subjects are not even aware of the true purpose of the test. To address this problematic issue, the committee recommended having medical personnel on hand to assure the “medical well being” of the subjects. Yet even those medical personnel would not be told the reason for the testing and would instead be given a cover story. Humanely, at least, the committee did recommend “gonadal protection be provided” to the male test subjects.

  Fortunately for the would-be recruits and their gonads, the human tests were never pursued. The committee’s views on Pandora quickly began to change as they reviewed the actual data, which eventually included more primates and additional testing. The scientific committee’s minutes, declassified and released years later, demonstrate increasing doubts about the testing protocol, in particular the lack of controls used in testing the monkeys. Among the concerns was that there was never a solid baseline established to compare how the monkeys’ performance allegedly degraded after exposure to radiation, the members noted. In other words, it was never established how well monkeys performed the tasks during a test period when not exposed to any periodic bouts of radiation.

  While Pandora never progressed to testing on humans, it did look at the effects of occupational radiation exposure on humans. One experimental protocol, called Big Boy, examined sailors on the USS Saratoga, comparing those who worked above deck, and were exposed to radiation from the radar, to those who worked below deck (the sailors were not told that they were part of a human radiation study; an unspecified cover story was used). The conclusion was that there were no psychological or physical effects as a result of exposure to low-level microwave radiation.

  In 1968, Joseph Sharp, the lead Pandora researcher at Walter Reed, left the program. Major James McIlwain, a medical doctor who had been drafted into the army, was selected to replace him. It took almost a year before McIlwain was cleared for Project Pandora, but once that happened, he got to work on a rigorous review of the data, poring over the computer printouts detailing each animal’s behavior. Within a year, McIlwain completed the statistical analysis, and what he found was not encouraging for the prospects of microwave mind control weapons. The basic question, he recalled in an interview years later, was whether it was more likely that the animal would stop working when the radiation was on compared with when it was off. “The answer to that was no,” he said. The Pandora scientific review committee agreed, concluding, “If there is an effect of the signal utilized to date on behavior and/or biological functions, it is too subtle or insignificant to be evident.” In other words, microwaves could not be used for mind control.

  By 1969, Stephen Lukasik, then the deputy director of ARPA, had some serious doubts about Cesaro, whom he regarded as a serial liar. The impresario of ARPA’s black programs acted as if he reported to no one, alluding to orders from high-level intelligence agencies but refusing to provide any specific information. “He was all over the place, cloaked in special access programs,” Lukasik said, a reference to h
ighly classified national security programs.

  Pandora, the mind control project, was particularly worrisome. At that point, the research had been going on for almost five years, and millions of dollars had been spent for construction of a new microwave laboratory. Lukasik asked Sam Koslov, the original director of ARPA’s Advanced Sensors Office, to review the Pandora file and let him know what he thought. Koslov was an old hand at intelligence projects and less likely to be snowed by claims of secrecy and overwrought concerns about the potential for Soviet exotic weapons. Koslov, then at Rand, reviewed the materials and discussed the results with McIlwain, at Walter Reed, and reported back to Lukasik in November 1969.

  Like other review committee members, Koslov criticized the original experiments for having almost no baseline and noted the experimental procedure changed over time. Also, if the question was whether a modulated microwave beam, such as the Moscow Signal, was causing deleterious effects, why was it never measured against a continuous wave? he asked. Simply zapping monkeys with the Moscow Signal was an entirely wrong approach, if the goal was to understand whether the effects were associated with a specific signal. “One should start with an examination of various basic wave forms and then the combinations resulting in possible intermodulations and demodulations by biological tissue,” Koslov wrote.

  Koslov also rightfully questioned whether the entire program truly needed to be secret. One could much better run a more open program that looked at the health effects of microwaves generally and then have a secret program looking at technology or weapons, if it was warranted, he argued. “In brief, I am forced to conclude that the data do not present any evidence of a behavioral change due to the presence of the special signal within the limits of any reasonable scientific criteria,” Koslov wrote to Lukasik. In 1969, ARPA ended its support for Pandora, and the remaining work was transferred to Walter Reed. At the time it was closed down, the program had spent upwards of $5 million—a sizable amount for a biological sciences program at the time and a good portion of the Advanced Sensors Office’s budget.

  Even before Pandora drew to a close, Cesaro found another area where he could market ARPA’s secret sensor technology: Vietnam. Instead of zapping people with microwaves, however, Cesaro was trying to use ARPA’s sensors to hunt them down and kill them.

  —

  In 1967, the American military in Vietnam was facing a politically delicate but militarily challenging situation: the North Vietnamese military was increasing its attacks across the demilitarized zone, but the United States was restricted by its rules of engagement, which required the military to identify what it was shooting at. “You can’t just hit a target because it moves, you have to know what it is,” Eberhardt Rechtin, ARPA’s director at the time, explained to lawmakers. “That means, in effect, that you have to be looking at it.”

  Cesaro, once again, had a solution, or many solutions, in fact. By the late 1960s, he had managed to expand his classified work from microwave research in the United States to a series of controversial Vietnam War projects. All involved using the sensor technology that ARPA had developed from the nuclear world and applying it to identifying targets. Cesaro was not always practical, but he was creative when it came to foreseeing the potential applications of technology. His most ambitious—and eventually most high-profile—plan was an audacious effort to arm surveillance drones in order to kill “high priority targets” in North Vietnam. Thirty-five years before the CIA used an armed Predator in Afghanistan, ARPA was preparing to arm a strange-looking drone that looked as if it were made out of spare parts in someone’s garage.

  When Constantine “Jack” Pappas, a young naval officer, showed up for a scheduled appointment at ARPA in 1967 to talk about ARPA’s drone plan, he got a quick lesson in why so many people disliked Richard Cesaro. After being kept waiting more than an hour for Cesaro, who was in his office, Pappas finally lost his temper at the secretary, threatening to “blow the door down” if Cesaro did not come out to see him. When Cesaro emerged, however, Pappas’s view soon changed. “Cesaro was a prickly guy with a big ego,” Pappas said, but he was genuinely interested in new technologies.

  Pappas had been assigned to the QH-50 Dash, short for Drone Anti-Submarine Helicopter, an unmanned aircraft armed with a nuclear weapon. The QH-50 was a coaxial helicopter, which means that it used two main rotors that spun in opposite directions, and had no tail rotor. The navy originally bought the drone helicopter with plans to use it in antisubmarine warfare, because its compact size made it ideal for operations off a small ship deck. The QH-50 would hunt for Soviet submarines; when it found them, it would drop a nuclear depth charge on them. The oddly shaped drone was innovative but prone to crashing, at least according to its critics. However, the QH-50 was secretly drafted into war in Vietnam by the U.S. Navy, which loaded the drone with sensors for reconnaissance. The sensor-laden version of the QH-50 was nicknamed Snoopy.

  In the fall of 1967, the air force and ARPA launched a project code-named Blow Hole, to find a way to deal with the North Vietnamese incursions across the demilitarized zone. The goal was to have a technology that could be delivered in forty-five days. ARPA’s plan was to take the helicopter drone, arm it, and send it flying over the DMZ in search of Vietcong, who were launching artillery at American forces. Cesaro initiated two drone projects in early 1968: a QH-50 loaded with a television camera and electronics, which was called Nite Panther; and a second, armed version, which was called Nite Gazelle. Over the course of the next four years, ARPA would experiment with putting guns, grenade launchers, bombs, and missiles on the QH-50. In one of the first demonstrations of “precision targeting,” the Nite Gazelle was equipped with a laser that designated targets, which could then be destroyed by weapons fired from air force or navy aircraft.

  In 1969, Cesaro expanded his killer drone work with a project called Egyptian Goose, which would place a radar on a “leftover” balloon from World War II (the name was derived from a parallel Israeli project that would use the balloons to spy on Egypt). Another balloon, called Grandview, would relay television surveillance video from the battlefield in Vietnam, identifying targets so that the armed drone could destroy them. Neither balloon made it to Vietnam, but the project was later credited with leading to the U.S. government’s use of tethered balloons for surveillance, which by the early twenty-first century ranged from protecting the American border with Mexico to military bases in Afghanistan.

  The QH-50, on the other hand, was sent to Vietnam, even though, as a later ARPA history noted, the aircraft had a “checkered” history, an understated way of referring to the drone’s multiple crashes. Reliability problems plagued the QH-50. Trying to get the weapons to accurately hit anything also proved futile. Shortly after the program started, the QH-50 equipped with a 7.62 mm mini-gun and gravity bombs was tested at the Naval Air Station in Patuxent River, Maryland. The tests, a later report notes, were “unsuccessful.” While some of ARPA’s QH-50s made it to Vietnam for testing, it appears none of them were ever used in combat. Stephen Lukasik says be believed all the armed QH-50s ended up crashing. Though Nite Gazelle was never used operationally as a weapon, it demonstrated the technological feasibility of a drone that could both spot and kill an enemy.

  Cesaro’s aggressive pushing of his office’s technology brought him into conflict with almost everyone, including Seymour Deitchman, who was in charge of AGILE at the time. Deitchman regarded Cesaro as an aggravation who pushed technologies without regard to their operational utility, like advanced night-vision devices that required an unwieldy battery pack. Deitchman recalled clashing with Cesaro over a project named Dancing Bells, which was supposed to spot Vietcong hiding in the underbrush. The idea was to equip helicopters with sensors that used constantly shifting frequencies to detect possible human movement under jungle foliage. Deitchman decided the vibration would make the images unusable and canceled the project over the objections of Cesaro.

  Cesaro’s office also paid for the MIT Lincoln Laboratory
to build radar that could see through the jungle, protecting military outposts from surprise Vietcong attacks. The Camp Sentinel Radar went to Vietnam in 1968 at Lai Khe. When it was mounted on top of a tall tower, its electromagnetic energy could penetrate the dense foliage to pick up any potential Vietcong lurking in the underbrush. Though the six prototypes of the radar sent to Vietnam were hailed as a technological success, Fred Wikner, a physicist who served as science adviser to General Creighton Abrams, said the only Sentinel radar he ever saw in Vietnam was knocked out during a typhoon and took two years to repair.

  —

  The legacy of ARPA’s most secretive and controversial office is mixed. By the end of the 1960s, the intelligence community concluded that the Soviets were using the pulsed radiation to activate listening bugs concealed in the embassy’s walls, and not to control diplomats’ minds. Yet concerns about the Moscow Signal lingered even after the scientific testing ended, though mind control was generally ruled out. A State Department doctor in charge of the blood tests, Cecil Jacobson, asserted that there had been some chromosomal changes, but none of the scientific reviews of his work seemed to back his view. Jacobson achieved infamy in later years, not for the Moscow Signal, but for fraud related to his fertility work. Among other misdeeds, he was sent to prison for impregnating possibly dozens of unsuspecting patients with his own sperm, rather than that of screened anonymous donors as they were expecting.