The Wizards of Langley

by Jeffrey T Richelson

  Bissell spelled out his objections to the proposed directorate. He questioned the wisdom of transferring the scientific intelligence (OSI) and pho-tointerpretation functions to the new directorate, thus “separating [them] from the other offices under the Deputy Director (Intelligence) that are concerned with the analysis and production of finished intelligence.” In addition, although he acknowledged that the DCI might need a policy adviser on signals intelligence issues, he had a “personal distaste for this role.” He further explained that if he had a deep interest in the signals intelligence field, he would prefer some operational or managerial position.11

  Bissell also questioned the idea of splitting the Technical Services Division, placing its research and development activities under the new directorate while leaving support functions with the DDP. He was “inclined to believe that progress in the exploitation of advanced techniques can be accelerated only by forcing a closer integration of developmental and operational activities which will be far easier to accomplish if they remain under common command.” As a result, he was not sure if any part of TSD should be transferred to the new directorate.12

  As for reconnaissance projects such as CORONA, OXCART, and SAMOS, he agreed that “responsibility for these special projects could well be placed elsewhere than in the Clandestine Service and that they would benefit from more top management attention than I have been able to give them for the past several years.” However, he questioned whether the CIA could expect to play a significant role in the future.13

  Even if the agency continued with OXCART and played subsidiary roles in CORONA and SAMOS, the CIA officer in charge of such activities, even if he also was responsible for some portion of TSD and other research and development, would more appropriately be an assistant to the DCI rather than a deputy. That position, Bissell wrote, “would have approximately the same scope as the one I occupied in this Agency in 1958. . . . For me to accept it would mean a long step backward.” Shortly afterward, Bissell sent a follow-up letter of resignation, effective February 17.14

  THE FOURTH DIRECTORATE

  Bissell’s resignation left McCone needing to find a manager for the CORONA, IDEALIST, and OXCART programs. It also led to renewed pressure from Killian and Land to establish a science and technology directorate removed from covert activities.15

  On February 14, McCone approved a “headquarters notice” announcing that Richard Helms would replace Bissell as DDP on the seventeenth and that plans were under way to establish a Deputy Directorate for Research and Development. Two days later, another headquarters notice informed readers that effective February 19, the agency would have a Deputy Director for Research (DD/R) at the head of a Deputy Directorate for Research (DDR); that certain functions of the DPD as well other research and development activities would be transferred to the new directorate “in the interest of strengthening the Agency’s technical and scientific capabilities”; and that Herbert “Pete” Scoville, longtime head of OSI, would head the new DDR. In late June, Col. Edward Giller, who had been serving as deputy director of TSD, became Scoville’s assistant deputy director.16

  Establishing an organizational structure for the new directorate was a prolonged process, in part because of continued opposition in other segments of the agency. On April 16, the transfer of the reconnaissance activities of the DPD’s Special Projects Branch to the DDR was authorized. The branch brought along responsibility for the CORONA, ARGON, IDEALIST, and OXCART programs.17

  But the “Battle of Charter Ridge,” as one Scoville deputy described it, continued into July. An early July memo to Scoville stated that “progress in defining [your] sphere of command and . . . functional responsibilities has been virtually negligible.” It suggested that Scoville might consider calling a halt, at least in the short term, to attempts to obtain the transfer of the Technical Services Division and Office of Scientific Intelligence to the new directorate and settle for control of the reconnaissance and ELINT functions.18

  Finally, in late July, a mission statement for the new organization was issued by Deputy DCI Marshall Carter. The new directorate was not all that had been promised. Its mission was “to conduct in-depth research and development in the scientific and technical fields to support intelligence by advanced technical means.” Excluded from its charter were research and development activities to support agent operations, as well as the non-ELINT duties of the Office of Communications.19

  Carter’s memo identified the three offices that would carry out the directorate’s responsibilities: The Office of Special Activities (OSA), established near the end of June, would manage the CIA’s reconnaissance programs. There would also be an Office of Research and Development (ORD) and an Office of ELINT (OEL), with the latter assuming responsibility for the ELINT activities that had been conducted by OSI and other CIA components.* The Plans directorate retained control over any clandestine agent operations or liaison activities involving ELINT as well as Division D’s embassy-based COMINT collection operations.20

  Two offices that were not assigned to the new directorate, and never would be during Scoville’s tenure, were his own Office of Scientific Intelligence and the Plans directorate’s Technical Services Division. Scov-ille had accepted the job on the basis of McCone’s promise to transfer the two organizations to the new directorate. But both Ray Cline and Richard Helms objected strenuously.21

  Cline, an Illinois-born, Harvard-educated veteran of the wartime Office of Strategic Services (OSS), had held both analytical and operational posts in the agency. He joined the Office of Reports and Estimates in 1949, headed the Estimates Staff of the Office of National Estimates (ONE), worked out of the London embassy from 1951 to 1953, transferred from ONE to the Office of Current Intelligence in 1955, and, late in 1957, accepted the position of Station Chief in Taiwan. In March 1962, McCone offered him the position of Deputy Director for Intelligence (DDI), to replace the departing Robert Amory. On April 23, Cline became DDI while disengaging himself from his work overseas.22

  Cline insisted that OSI remain in his directorate, arguing that removing it would mean “weakening the CIA’s analytical voice.” He contended that all the analytical units should remain part of the same organization, and that collection and analysis should be handled by distinct components. Richard Helms, the new DDP, felt equally strongly about keeping the essence of TSD, along with some of the directorate’s aircraft operations that involved support to covert activities, in his directorate.23

  There were a number of reasons for Helms’s reluctance to part with TSD. A memo noted his “alarm” at any encroachment of his authority to conduct overseas clandestine operations, the concern that the gulf that already existed between those developing agent equipment and those using it would be exacerbated by TSD’s transfer out of the operations directorate, and the aversion of Plans to provide detailed information on clandestine operations to technicians in another directorate.24

  Of course, as much as Cline and Helms wished to retain control of OSI and TSD, respectively, a mere stroke of a pen by McCone would have effected the transfers. But despite, or perhaps because of, Cline and Helms being McCone appointees, the DCI refused to deliver OSI and TSD to Scoville. Scoville soon discovered that holding McCone to his commitments would be a problem—that the steely-eyed, rock-ribbed Republican was not always as tough as he looked.

  ORD

  While the OSA and OEL groups were created from existing components of the CIA, the Office of Research and Development (ORD) was a creation of Scoville and his deputy, Air Force Colonel Edward Giller, who also became its first head. They envisioned an office that would look several years into the future, investing in research and development activities that might pay dividends for collectors or analysts in five or ten years.25 Indeed, failure was part of ORD’s mission over the years, not only to find out what would work but also what wouldn’t—before another part of the agency poured millions of dollars into a doomed effort.

  ORD began work in January 1963 with a st
aff of three individuals transferred from the technical services division—one person for each of its divisions (Research, Systems, Analysis). The initial focus of the office was to be “research and development . . . to support intelligence collection by advanced technical means.” Topics of particular interest identified in ORD’s charter were new optical systems leading to improved resolution; use of lasers to permit night photography; chemistry research related to collection concerning biological and chemical warfare activities; and acoustic and seismic research related to missile intelligence.26

  In addition to research, ORD (through its Systems Division) was to analyze promising ideas falling outside the responsibility of the directorate’s overhead and ELINT collection offices and turn them into technical collection systems. The division even fielded collection systems and assumed responsibility for the EARTHLING site in Pakistan.27

  Part of ORD’s initial charter was to assume TSD’s main research functions, including in behavioral science, leaving that organization to handle the operational support and related R&D functions that Helms believed must remain in Plans. Thus, ORD took over part of the MKULTRA program. Dr. Stephen Aldrich, a graduate of Amherst and Northwestern Medical School who had served in the agency’s Office of Medical Services and OSI’s Life Sciences Division, assumed many of the responsibilities that had belonged to Sidney Gottlieb.28

  With Aldrich directing that portion of ORD’s activities, its scientists continued searching for ways of controlling human behavior. The research they inherited from TSD included placing electrodes in the brains of dogs and other animals and then using radio signals to guide them along specific courses. The technical services division also placed electrodes into the brains of cold-blooded animals—apparently snakes. The experiments with dogs were directed at bugging an office, but the experiments with cold-blooded animals may have had a more cold-blooded objective, possibly “executive action-type assassinations.”29

  Creating amnesia remained a major objective of ORD. Advances in brain surgery facilitated far simpler psychosurgery and the possibility that “a precisely located electrode probe could be used to cut the link between past memory and current recall.” According to one account, ORD had access to prisoners in at least one American penal institution, and office staffers worked with the Edgewood chemical laboratory to develop a drug that could be used to implant false memories into the mind of an amnesia subject.30

  ORD also supported work done at the Scientific Engineering Institute (SEI). A group of behavioral and medical scientists was permitted to conduct independent research as long as it met SEI standards. The scientists were available to consult with frequent visitors from Washington. One project apparently involved stimulating the pleasure centers of crows’ brains in order to control the birds’ behavior.31

  ELINT

  Heading the new Office of ELINT (OEL) was George Miller, who had been serving as head of OSI’s electronic intelligence effort. The activities he managed as OEL chief included support to the Norwegian Kirkenes facility, the Norwegian Barents Sea boat operation, the CIA ELINT collectors in Iran, and the “Quality ELINT” program. Support for an Austrian COMINT station, essentially on behalf of NSA, was also a mission of the office.

  Collection targets continued to include Soviet space activities as well as early warning radars that were in the developmental stage. On April 26, 1962, the Soviet Union launched Cosmos 3, identified only as “a new data transmission system,” which was actually a test of a SAMOS-type photographic system (code-named BAIKAL), employing electronic readout to return the data to earth. CIA ELINT specialists attempted to demodulate the signals and similar ones from the Cosmos launch of July 28. They were able to establish some of the signal parameters but could not produce identifiable pictures. They did conclude that one or more cameras on board were taking photos, probably of cloud formations, and that the film was developed by an on-board processor. In addition, they determined that the film was being electronically scanned and transmitted to ground-based receivers in the Soviet Union. The effort directed against Cosmos 9, launched on September 27, which carried the same type of system as Cosmos 3, was more successful. Cloud cover was easily identified in a series of six pictures, and further analysis led to the conclusion that Cosmos 9 was an experimental weather satellite.32

  ELINT collection and analysis also focused on two targets that appeared in 1960 U-2 and CORONA photography—a pair of radars the U.S. intelligence community designated Hen House and Hen Roost. Both were located on the western shore of Lake Balkhash in the USSR and looked out from the Sary Shagan antiballistic missile (ABM) test center toward Kapustin Yar, the launch point for ballistic missiles employed in ABM tests. Both facilities were enormous. The Hen House antenna building was over 900 feet long—more than 3 football fields—and nearly 50 feet high. The Hen Roost radar had 2 antennae, a half-mile apart, each over 500 feet long. One antenna was a mere 15 feet tall; the other reached 65 feet.33

  Radars of this type were just being developed in the United States. Rather than employing traditional radar dishes that were mechanically steered, the face of these phased-array radars consisted of radiating elements. The delay in the signals sent out from those elements meant a beam could be electronically steered to detect incoming objects. But the CIA, the Strategic Air Command, and other elements of the national security establishment needed to know more. They needed information on the system’s operating characteristics in order to determine whether the radars could provide the data required by Soviet ABMs to destroy incoming U.S. warheads, and how to neutralize the radars. But electronic monitoring of any activity out of Sary Shagan was virtually impossible at the time. The site was over the horizon from all U.S. ground stations, and the electronic signals emitted by Hen House or Hen Roost would head off into space before they could be intercepted by U.S. antennae.

  The first inroad came in late 1962, due to the Soviet decision to renew atmospheric testing of nuclear weapons, testing that involved detonating missile warheads as they approached Sary Shagan. Two missiles were fired sequentially from Kapustin Yar; the warhead of the first was detonated in the atmosphere to determine if the Sary Shagan radars could detect the second missile through the nuclear cloud. The release of such vast amounts of energy can cause radical changes in the radio transmission properties of the surrounding atmosphere. In the highly ionized region created by a nuclear blast, radar waves, which would ordinarily travel straight into space, can be reflected or bent in different directions. On October 28, one of the ABM-related tests at Sary Shagan had just that effect. U.S. ELINT stations in the Middle East, possibly including Beshahr, recorded thirteen new signals, many of which were believed to have originated in the Sary Shagan region.34

  The signal of greatest interest was originally designated BUEB, which analysis indicated was designed to be used against targets more than 800 miles away. Aircraft flying at the highest altitudes are well below the horizon when they are 300–400 miles from a radar, and they travel rather slowly. Ballistic missiles rise several hundred miles above the earth and approach their target rapidly, making it desirable to detect them as far away as possible. BUEB therefore became a prime candidate to be an ABM radar signal. In addition, each pulse was transmitted at a different frequency, which would be expected from an electronically steered radar beam. One major question remained: Which radar was the United States hearing, Hen House or Hen Roost? It would be a few more years before that question would be settled.35

  MELODY AND PALLADIUM

  Soviet early warning radars were only one focus of the CIA’s ELINT effort. Another was the air defense radars that would be crucial to any attempt to detect, track, and destroy U.S. reconnaissance flights that might overfly Soviet territory. That effort had begun in 1959—before the May 1960 U-2 shootdown and while the OXCART program was in its early stages. OXCART staffers wanted to know how widespread the radars were, the extent of territory they covered, the power they radiated, and their sensitivity. While seeking to answ
er the first question, the ELINT unit of OSI discovered a technique that enabled the Beshahr station in Iran to monitor Soviet missile tracking radars and eliminated the need to rely on chance occurrences such as that of late October 1962.

  Gene Poteat, a member of the ELINT staff, joined the CIA after working as an electrical engineer and physicist for Bell Labs, reporting to work in early January 1960. He recalled an occasion at Cape Canaveral in the 1950s when a signal was received from a ground-based radar located 1,000 miles beyond the horizon because the signal had reflected off a Thor IRBM (Intermediate-Range Ballistic Missile) during a test flight. This event led to the suggestion that the phenomenon, which would later become known as “bistatic intercept,” could be used to locate high-powered Soviet radars over the horizon from U.S. intercept sites. Instead of pointing ELINT antennae in the direction of possible Soviet radars, the antennae would seek to capture signals reflected off the missiles.36

  George Miller suggested Poteat try out the idea on two experts from private industry—William Perry of Sylvania and Albert “Bud” Wheelon of TRW. Both men offered moral and technical support. Feasibility studies and engineering calculations followed, which required solving spherical trigonometry equations using slide rules, logarithm tables, and hand-operated mechanical calculators.37

  Funding came quickly for the program, which Poteat code-named MELODY after one of his favorite sounding words, and the appropriate equipment was installed at the Beshahr site. The MELODY program produced bistatic intercepts of virtually all Soviet ground-based tracking radars, including all of the ABM radars at the Sary Shagan test range.38