Inside the Department of Dirty Tricks

7/24/2019 Inside the Department of Dirty Tricks

1/49

As originally published in The Atlantic Monthly,August 1979

"We're not in the BoyScouts," Richard Helmswas fond of saying whenhe ran the Central

Intelligence Agency. Hewas correct, of course. Boy

Scouts do not ordinarilybribe foreign politicians,invade other countries withsecret armies, spread lies,conduct medicalexperiments, build stocksof poison, pass machine

guns to people who plan toturn them on their leaders, or plot to kill men such as

Lumumba or Castro or others who displeased

Washington. The CIA did these things, and more, over along span of years. On whose orders? This is a question aPulitzer prizewinning writer addresses in an adaptationfrom his forthcoming book about Helms and the Agency,The Man Who Kept the Secrets.

by Thomas Powers

1. An Isolated Man

RICHARD Helms, as lean as a long-distance runner andlooking just about as restless, dressed in a suit and tie,greeted a visitor at nine o'clock on a sunny morning on hisfront doorstep. He would not have been dressed anydifferently if he'd been on his way to present an annualreport to the board of directors, but in the spring of 1977he was not going anywhere. The reason was not that hewas looking forward to a chance at last to read thecollected novels of Balzac, or that he wanted to stay hometo work on his stamp collection, or that he welcomed the


2/49

freedom to watch a whole season of baseball on television.The reason was that his whole life was hanging fire whilehe waited to learn if a special grand jury in the District ofColumbia would vote to indict him for certain actscommitted shortly after he ceased to be director of centralintelligence (DCI) of the CIA.

Return toFlashback:Department ofDirty Tricks

Indicted for what? Helms would ask in his own defense.Helms is a man with an oddly appealing grin. His lower

jaw juts out a trace, giving his otherwise ordinarilyhandsome face a singularity. His grin, lower jaw out, eyeswide, hands up, has about it an ironic, incredulous air; hecan be amused, bewildered, and angry at the same time.For what?

He knew perfectly well for what, but intended to conveyhis contention that he had never done anything he was notasked, ordered, expected, or required to do by the nature ofhis job. In particular, the director of central intelligence

had a responsibility not to answer every idle question putto him. He was charged under the National Security Act of1947 with protection of the CIA's sources and methods. Noone has ever spelled out what powers are thereby grantedto the DCI. Helms had to protect the CIA's secrets byhimself. It was his job and he did it. Indicted for what?

The narrow answer was for perjury before the SenateForeign Relations Committee on February 7, 1973, whenHelms answered a question put by Senator StuartSymington -- "Did you try in the Central Intelligence

Agency to overthrow the government in Chile?" -- with anunequivocal "No, sir."

"Did you have any money passed to the opponents ofAllende?"

"No, sir."

"So the stories you were involved in that war are wrong?"

"Yes, sir."

Helms's problems added up to a general mess of a sortunthinkable in previous years. But the dimensions andpossible consequences of the mess had not yet halted theinvestigation, despite quiet appeals to the JusticeDepartment by distinguished Washington figures whothought Helms was getting a raw deal. Taken together,these facts explained why Helms, who ought to haveencountered little difficulty in finding a job, was not free towrite his memoirs or accept employment or do much ofanything except play tennis, dine with friends, and wait for
http://www.theatlantic.com/past/unbound/flashbks/cia/cia.htm


3/49

his lawyer to straighten things out.

Helms was an isolated man. It was not that he lackedfriends and allies in Washington, where he had spentnearly thirty years in the practice of intelligence. He wasboth liked and respected there, on his chosen ground; he

was taken to be an honest man, a dedicated public servantwho deserved honorable retirement after a long careerworking his way up through the ranks of the CentralIntelligence Agency. Not many people knew what Helmshad been doing in the CIA, but those who did formed acircle of unusual power and influence -- former Presidents,cabinet secretaries, and other high officials, congressmen,and leading journalists. But this phalanx of support,personally gratifying as it must have been, onlyemphasized his isolation. Outside Washington, the word"intelligence" had acquired a new and sinister shade of

meaning. Four years of official investigations had cast theCIA in a dark light, and the name of Richard Helms hadturned up on a great many embarrassing documents aboutWatergate intrigue, assassination plots, the testing of drugson unwitting victims, attempts to foment coups indemocratic countries. The Washington circle that excusedthese things, explaining them away as the prosaic facts ofinternational life, was a decidedly small one, and Helmswas trapped at its very heart.

Helms did not understand how this had happened. He

certainly knew the details of recent history better thanmost. He had watched the awful progress of events fromWatergate to a major investigation of the CIA by a Senateselect committee, and he had resisted the process ofexposure at every step of the way. Helms had feared twoconsequences from the hemorrhage of Agency secretswhich was still continuing: the demoralization of the CIA,unaccustomed to public scrutiny and a field day for hostileintelligence services rummaging through the Senatecommittee's voluminous reports. In Helms's view both had

occurred, just as predicted. He was not a believer incatharsis. He was neither embarrassed nor repentant. Menof the world knew that the business of intelligence wasmore than a simple matter of spy and counterspy. WhatHelms did not understand was the relentless harping --especially on the part of certain Senate liberals and thepressmen -- on the "crimes" of the CIA. Of course Helmsread the papers; he knew there was a large public that didnot like the Agency and what it was taken to represent --the secret expedients of power, and the failures ofAmerican Presidents who had tried to bull their way in the


4/49

world. The wreckage of Vietnam was proof enough thatsomething had gone terribly wrong. But in Helms's view,the hostility focused on the Agency, and indirectly on him,was the result of a refusal to accept the reality of ananarchic international system, in which vigilance, power,and strength of will were a nation's best, indeed only,defense. Destruction of the CIA through exposure andrecrimination was like spiking the guns.

In the spring of 1977, out of a job for the first time innearly forty years, Helms had plenty of time on his hands;his lawyer had told him to keep out of the public eye. Butit went against the grain. Temperament and years of habithad accustomed him to days of busy executive routine:office by 8:30, meetings throughout the day, the review ofendless pieces of paper, departure regularly at 6:30. CIApeople like to tell stories about the Agency's great days andthe adventurous men who ran its operations beforeeverything fell apart, but they do not tell anecdotes about

Helms: there aren't any. He is remembered as anadministrator, impatient with delay, excuses, self-seeking,and the sour air of office politics. Asked for an example ofHelms's characteristic utterance, three of his old friendscame up with the same dry phrase: "Let's get on with it."He had hired out to do a job, he did today what had to bedone today, he left his desk clean at night.

Of course every desk at the CIA was clean at night. Thesecurity people roamed the building after the close of workand handed out demerits for unlocked safes, full trashbaskets, classified documents left in desk drawer. Even thedesk of a man such as Richard Bissell, Helms'spredecessor as head of the CIA's Deputy Directorate forPlans (DDP), had been clean at night before he left theAgency in disgrace after the collapse of his plan to invadeCuba at the Bay of Pigs. It would be hard to imagine twomen more unlike than Richard Helms and Richard Bissell.Helms had been pretty unhappy when Bissell got the jobHelms wanted back in 1958, but it wasn't solely personaldisappointment that distressed him. Bissell was

loquacious, inventive, the most open-minded of men; therewas literally nothing one might propose to him that hewould not turn this way and that in his logical mind,

judging it strictly on its practical merits. A plan to invadeCuba, a poisoned handkerchief for an Arab general -- hewas ready to entertain them all. But Bissell's logical claritywas illusory. He sometimes fatally misjudged men. Heworked out schemes for management and then broke hisown rules. His desk was chaos. One look at it (and Helmsdid not get many; Bissell did not invite Helms's advice)and one might despair for the country. But even Richard


5/49

Bissell's desk, straightened up by his secretary, was cleanat the end of the day.

No branch of the American government was in better orderat night than the CIA, in its huge headquarters in themiddle of a woods in Langley, Virginia. It was the biggestthing of its kind in the world, much larger and moremodern than the headquarters of the Committee for State

Security -- the KGB -- in Moscow. The nation's secretswere each in their appointed place and one might havethought, if one had made the rounds with the securityofficers checking for violations, that the country must be ingood order, that everyone knew his job, and accepted theground rules, and agreed on the importance and purpose ofthe business at hand. An illusion, as Richard Bissellabruptly discovered in April 1961.

Helms had not been much surprised by Bissell's failure atthe time. But he cannot have imagined, as he picked up the

pieces in Bissell's wake, that his own gifts as anadministrator, his long experience in managing secretoperations, his devotion to their secrecy, his caution andcool judgment, would all fail too. Indeed, before hisgovernment was through with him, Helms would havereason to envy Bissell's quiet departure. The problem wasnot the way Helms or Bissell or anyone else in the CIAhad been going about his job, but the job itself. Theproblem was what they did. The meticulous routine andorder of the Agency, the tables of organization, the well-established and accepted dealings with the other branchesof government, the procedures for internal and externalcontrol, the apparent consensus of official Washington onthe importance of the CIA's work, were all illusory. Thestructure was jerry-built. The agreement was mostlyconfined to a small circle in Washington.

The arrangement had worked so well for so long that itwas hard to see how fragile it was. The foreign policyestablishment in Washington trusted the CIA, and stilltrusts it, for that matter; but beyond governing circles the

political foundation of the CIA rested on nothing moresubstantial than a popular fascination with espionage and aconviction that we are the good guys. The Americanpublic, in short, had been taught a kind of child's history ofthe world, sanitized of the rougher facts of internationallife. A Victorian political morality obtained. Presidents,congressional leaders, the Pentagon, and the StateDepartment all found it convenient to let the public assumethat only the Other Side did things like that. We did notbribe foreign politicians. We did not undermine othergovernments. We did not invade other countries with


6/49

secret armies. We did not spread lies, conduct medicalexperiments, put prisoners in padded rooms for years onend, build stocks of poison, sabotage factories,contaminate foodstuffs, pass machine guns to men whoplanned to turn them on their national leaders. Above all,we did not plot to kill men for nothing more thandispleasing Washington. To discover ourselves the victimsof so many illusions, all at once, was disorienting. The

result has been a profound shift in public attitudes anddeep confusion in Washington, where simultaneous effortsare under way to make sure the Victorian morality reallyobtains this time; to deny that it was ever seriouslybreached; and to get the CIA back on the job.

2. The Regular Spiel

THE business of intelligence has its ugly side. Theimmaculate documents that go to the National SecurityCouncil do not come only from satellites and a closereading of Russian technical journals. Presidents haveways of getting their message across which go beyondState Department white papers and speeches in the UN.Secret agents must be not only recruited but controlled.When they go sour they may be betrayed to their enemies.Clients are sometimes led out onto limbs and abandonedthere. Allies of convenience are sometimes addicted to nail

pliers and electric needles. Friendly intelligence services,trained by the CIA in computerized file-keeping,sometimes use those computers to pull the names of peoplethey intend to kill. Helms spent thirty years in this businessand accepted it. "We're not in the Boy Scouts," he oftensaid. "If we'd wanted to be in the Boy Scouts, we'd have

joined the Boy Scouts." But the evidence, fragmentary as itis, suggests that the CIA customarily drew the line at whatis commonly meant by the word "murder." However, inthe late 1950s, the CIA began to get orders to kill people.

Of course talk about killing was a commonplace. In 1952,a West German general had lunch in Washington withDulles, Helms, and other CIA officials, and suggested thata way be found to assassinate the East German leader,Walter Ulbricht. The proposal was rejected. As early as1957, some American government officials were talkingabout "getting rid of" Ngo Dinh Diem in South Vietnam.His abuse of human rights was putting the United States inan awkward situation, just as his suppression of political


7/49

opponents of every stripe was undermining his owngovernment, but he had settled in so deeply that some ofthe Americans talking about getting rid of him had decidedthe only way was to get rid of him. Eventually thesediscussions involved the CIA; Vietnam analysts wereasked to suggest a possible replacement. In late 1958, notlong before he left the Agency for good, Frank Wisner,head of the DDP, discussed the "Diem problem" with

another DDP official, who says neither of them wasexactly keen on the idea. Diem with all his faults was anAmerican ally and client. "Is it really our job to do that?"the DDP official asked Wisner.

The answer turned out to be no -- for the time being atleast -- because the only man with a right to issue such anorder never gave it. But there was plenty of tough talk allthe same. At a State Department meeting to discuss U.S.troubles with Gamal Abdel Nasser of Egypt in late 1956 orearly 1957, Allen Dulles, suddenly growing angry with a

briefer's attempt to explain the situation from Nasser'spoint of view, turned to him and said, "If that colonel ofyours pushes us too far, we will break him in half!"

Later, in the 1960s, a member of the President's ForeignIntelligence Advisory Board, Robert Murphy, asked whythe CIA didn't kill Ho Chi Minh, since he was giving us somuch trouble; asked loudly, positively and repeatedly: Hois the problem, isn't he? Can't you fellows do something toget rid of him? You're supposed to be able to handlethings; handle him! Murphy was an important publicofficial and Thomas Karamessines, then head of the DDP,had a hard time with his repeated demands. A CIA officerwho often accompanied Karamessines said he'd heard suchtough talk before, and that he and other CIA officersresponded with "a regular spiel you'd give these people":What good would it do? Ho's successor might be evenworse. How were you going to kill Ho Chi Minh secretly?You might be able to fool the New York Times, RobertMurphy was told, but how were you going to deceive theVietnamese? They'd know what had happened, they'd

know who did it, and they'd probably be in a position andmood to retaliate. There is a tacit truce between nations onsuch matters once you start killing their people, they startkilling yours. The CIA simply does not have the assets tokill secretly a well-guarded figure such as Ho Chi Minh ina security-conscious state such as North Vietnam.

Another government figure who got the regular spiel wasLivingston Merchant, the undersecretary of state forpolitical affairs at the end of Eisenhower's Administration.Because of his job, Merchant was a regular member of the


8/49

Special Group to oversee covert operations, and onNovember 3, 1960, when planning for the Bay of Pigs waswell under way, Merchant attended a Special Groupmeeting where he asked "whether any real planning hadbeen done for taking direct positive action against Fidel,Raul, and Che Guevara." In the sudden absence of allthree, Merchant suggested, the Cuban government wouldbe "leaderless and probably brainless."

On this occasion the regular spiel came from GeneralCharles Cabell, the deputy DCI, who "pointed out thataction of this kind is uncertain of results and highlydangerous in conception and execution, because theinstruments must be Cubans. [Cabell] felt that, particularlybecause of the necessity of simultaneous action, it shouldhave to be concluded that Mr. Merchant's suggestion isbeyond our capabilities." Even through the opacity ofofficial minutes the pattern is apparent: a hardheaded,straightforward question -- What about it? If we're trying

to get rid of these guys, why don't we get rid of theseguys? -- is met with a wall of spongy demurrer: It's tootough, won't work, can't predict the consequences, mightblow up in our faces, et cetera.

A witness to still another episode of the sort was ArminMeyer, a career diplomat with a long history in the NearEast. In July 1958, when the government of Iraq wasoverthrown in a coup notable for its violence, Meyer wasdeputy director of the State Department's Office of NearEastern Affairs. The following year he was promoted todirector, and in that capacity was called in whenever theCIA contemplated covert operations in Iraq. The new rulerof the country was an army general named Abdul KarimKassem, who had murdered his predecessors as well as anumber of foreigners who happened to be in Baghdad atthe time of his coup. On top of that he immediatelyrestored diplomatic relations with the Soviet Union, laterlifted a ban on the Iraqi Communist party whilesuppressing pro-Western parties, and in many other waysinvited the hostility of Eisenhower and John Foster Dulles.

On one occasion during Armin Meyer's tenure as directorof the Office of Near Eastern Affairs, he attended ameeting in Allen Dulles's office at the CIA to discuss howthe United States might remove Kassem. Meyer hadattended many such meetings; they were a routine ofgovernment; but this one in particular stuck in his mind.

During the meeting one of those present suggested thatKassem himself was the problem, and maybe the best wayto get rid of him was to get rid of him. Wait a minuteDulles said. An awful silence followed. Dulles was a man


9/49

of great personal authority, and his words on this occasionhad a cold and deliberate emphasis which Meyer neverforgot. Dulles wanted one thing to be understood: it wasnot in the American character to assassinate opponents;murder was not to be discussed in his office, now or everagain; he did not ever want to hear another such suggestionby a servant of the United States government; that was notthe way Americans did things.

Dulles was so clear on this point, and spoke with suchevident passion and conviction, that Meyer, later, simplycould not understand how Dulles ever could have beenparty to an assassination plot, no matter who gave theorders. Meyer knew what was in the reports of the SenateSelect Committee on Intelligence (the Church Committee),but he simply did not believe it. Dulles had left no roomfor doubt: he would not be a party to assassination. Theregular spiel. The more one examines the subject, theclearer the pattern becomes. Another example ought to

make it unmistakable. On August 10, 1962, during theearliest stages of what would shortly become the Cubanmissile crisis, a meeting was held in the office of Secretaryof State Dean Rusk to discuss Operation Mongoose, theKennedy Administration's post-Bay of Pigs plan to get ridof Castro, and Secretary of Defense Robert McNamara, aman convinced there is a rational solution to everyproblem, was probably astonished at the instantaneousreaction to his entirely hypothetical suggestion thatperhaps they ought to consider solving the Castro problemby killing him.

Edward R. Murrow, the director of the United StatesInformation Agency, protested that this was entirely out oforder. CIA Director John McCone immediately backedhim up. The secretary at the meeting, Thomas Parrott, didnot so much as include the matter in the minutes. To sealthe point, McCone personally phoned McNamara later inthe day and protested that talk of assassination was entirelyout of order in such a meeting, that he didn't want to hearany more of it, and that he, McCone, a devout Catholic

who attended mass every morning, might be faced withnothing short of excommunication if word of such thingsever got out.

The message to McNamara ought to have been loud andclear: assassination was too sensitive a matter to bediscussed in official meetings or to be recorded in officialmemos and minutes. What those high officials whoreceived the regular spiel failed to comprehend was thedegree of secrecy which necessarily surrounded any matteras explosive as assassination.


10/49

In February 1960, while the government was trying todecide what to do about General Kassem, the chief of theDDP's Near East Division, James Critchfield, proposedthat Kassem be "incapacitated" with a poisonedhandkerchief prepared by the DDP's Technical ServicesDivision. In April the proposal was supported by theDDP's chief of operations, Richard Helms, who endorsedKassem's incapacitation as "highly desirable." As head of

the DDP, Bissell did not act in such matters withoutDulles's approval, and Bissell was convinced -- he couldhardly have made this point any clearer in his latertestimony before the Church Committee -- that Dulleswould not have proceeded without an order from the onlyman with the authority to okay an attempt on a foreignleader's life.

In this instance the handkerchief was duly dispatched toKassem, but whether or not it ever reached him, itcertainly did not kill him. His countrymen did that on

February 8, 1963, by machine-gunning him and three ofhis aides in his office in Baghdad.

What Livingston Merchant, Armin Meyer, RobertMcNamara, and others failed to understand was thatofficial meetings in the office of the director of the CIA, orof the secretary of state, or of the Special Group, werehardly the place to discuss something that was reallysecret. From the CIA's point of view, the secretary ofstate's office was about as secure as the floor of Congresswith full press galleries. If you were going to plan anassassination in the secretary of state's office, or record thediscussion in the minutes, you might as well send a pressrelease to the New York Times. Eisenhower and Kennedywent after two enemies in particular in the years between1959 and 1963 -- Lumumba in the Congo and Castro inCuba -- but when they gave the job to the CIA, theyexpected secrecy, and that is what they got.

3. A Case History: Cuba

THE Bay of Pigs marked the beginning, not the end, ofJohn F. Kennedy's determination to get rid of Castro, themoment when Fidel Castro ceased to be merely an enemyinherited from Eisenhower. Kennedy's mandate to GeneralMaxwell Taylor in April 1961 was, not to fix the blame forthe failure of the invasion, but to find out why it hadn'tworked, so the next plan would.


11/49

Knowing of Kennedy's growing obsession withunconventional warfare, Taylor proposed a broad,government-wide effort to combat insurgencies fromVietnam to Latin America. The result, after Taylor joinedthe White House full time as the military representative ofthe President on July 4, 1961, was establishment of theCounter-Insurgency (CI) Group, which began to meet on aregular basis with Taylor as chairman early that fall.

The first order of business for the CI Group was Cuba. TheCIA was heavily involved in both Laos and Vietnam at thesame time, but the covert operations launched againstNorth Vietnam, beginning in the fall of 1961 under theSaigon station chief, William Colby, were on the backburner. Cuba was where the Kennedys wanted immediateresults. A second committee, the Special GroupAugmented (SGA), was established to oversee OperationMongoose, run by then Colonel Edward G. Lansdale, acounterinsurgency specialist with experience in both the

Philippines and Vietnam, where he had helped Ngo DinhDiem to consolidate his control over the country. NoKennedy program received less publicity than Mongoose,or more personal attention from the Kennedys, and inparticular from Robert.

The importance of the undertaking did not take long toestablish. In the early stages of Mongoose, a CIA officerworking on the operation, Sam Halpern, asked LawrenceHouston, the CIA's general counsel, if the operation waseven legal. He pointed out that the Bay of Pigs landing hadbeen organized outside the United States at least partly inorder to avoid violating the Neutrality Acts, whichprohibited the launching of attacks on foreign targets fromAmerican soil. Now Mongoose was being geared up inMiami; wasn't this against the law? Houston said theanswer was no: if the President says it's okay, and if theattorney general says it's okay, then it's okay.

The CIA officers in charge of the Cuban branch set up byHelms were appalled by the magnitude of the task. "With

what?" they asked. "We haven't got any assets. We don'teven know what's going on in Cuba."

Despite the White House pressure, the SGA and Mongooseproceeded sluggishly. Lansdale's original plan had calledfor an escalating effort to create an opposition to Castroinside Cuba, followed by insurgency and a generaluprising. Lansdale spoke of a march on Havana in October1962, and he meant march -- a triumphal entry likeCastro's just three years earlier. But Lansdale's plan was afantasy. The CIA managed to get agents onto the island


12/49

and to recruit others in rural areas, but what they toldLansdale was bleak: there would be no general uprising.

After the first few months of covert operations, Mongoosegradually shifted its emphasis from resistance-building tosabotage, paramilitary raids, efforts to disrupt the Cubaneconomy by contaminating sugar exports, circulatingcounterfeit money and ration books, and the like. "We

want boom and bang on the island," Lansdale said. RobertKennedy took a particular interest in efforts to sabotage theMatahambre copper mines in western Cuba, on oneoccasion even calling repeatedly to learn if the agents hadleft yet Had they landed? Had they reached the mines?Had they destroyed them successfully? Kennedy, likeLansdale, wanted boom and bang, and a number of CIAofficers on the operational level grew to know his voice ashe called to find out how they were coming along and topress them forward. The Matahambre copper mines werenever destroyed, despite the launching of three separate

full-scale raids, but other attacks on sugar refineries, oilstorage facilities, and the like were more successful. Still,they fell far short of wrecking the Cuban economy, even inits weakened state following the dislocations of revolution,and the paramilitary program held little promise of Castro'soverthrow.

There is a certain opaque quality to all of the CIA's plansto eliminate Castro. The invasion force that landed at theBay of Pigs was too big to hide and too small to defeatCastro's huge army and militia. Mongoose in 1962 nevergot much beyond an intelligence-gathering effort, andwhile it succeeded in raising the level of "boom and bangon the island" in 1963, noise was hardly enough to do the

job. Lansdale's scenario for a triumphal march into Havanawas illusory. Desmond FitzGerald took over in 1963, but alot of people who worked for FitzGerald never quitegrasped how his plans were supposed to work either.FitzGerald was adamant. "You don't know what you'retalking about," he told one of them. They were going to getCastro.

But Lee Harvey Oswald got Kennedy first. After thePresident's murder in Dallas on November 22, 1963, theCuban operation began to wither away. The last exilegroups, boats, and maintenance facilities in Florida werenot abandoned until 1965, but Lyndon Johnson never gavehis full attention to the Cuban "problem."

In March 1964, Desmond FitzGerald, by then the newWestern Hemisphere division chief, visited the CIA stationin Buenos Aires. There he told some of his officers, "If


13/49

Jack Kennedy had lived, I can assure you we would havegotten rid of Castro by last Christmas. Unfortunately, thenew President isn't as gung-ho on fighting Castro asKennedy was.

Christmas, 1963. What could have "gotten rid of Castro"by Christmas of 1963?

There was a lot of talk about language at the meetings ofthe Church Committee. CIA officers testified that phrasessuch as "getting rid of Castro" were only figures of speech;they just wanted him out of the way, not dead and buried.It was a kind of shorthand, reflecting the determined spiritof the time. Perhaps they talked about "eliminating"Castro, or even "knocking him off," but they intended onlyto replace or remove him, not literally get rid of him. Ahandful of former CIA officials -- notably Richard Bissell,

William Harvey, Justin O'Donnell, Richard Helms --admitted that talk of getting rid of Castro or Lumumbameant just that in one or two instances, but when theyreally meant "get rid of," they sometimes used acircumlocution or euphemism instead. In particular, theytestified, conversations with high government officials,and especially any that might have occurred with the veryhighest government official, were deliberately opaque,Allusive, and indirect, using "rather general terms," inBissell's phrase.

In his Church Committee testimony, Helms took Bissell'sline. "I think any of us would have found it very difficultto discuss assassinations with a President of the U.S.,"Helms told the committee. "I just think we all had thefeeling that we're hired out to keep those things out of theOval Office." He made this point repeatedly -- "Nobodywants to embarrass a President of the United States bydiscussing the assassination of foreign leaders in hispresence"; "I don't see how one would have expected that athing like killing or murdering or assassinating would

become a part of a large group of people sitting around atable in the United States government"; "I don't knowwhether it was in training, experience, tradition, or exactlywhat one points to, but I think to go up to a Cabinet officerand say, Am I right in assuming that you want me toassassinate Castro? ... is a question it wouldn't haveoccurred to me to ask." Bissell and Helms both insistedthey had never discussed the assassination plots with eitherthe President or the attorney general, but at the same timethey were certain they had all the authority they needed,and were in fact trying to do what the Kennedys in


14/49

particular wanted done. Helms insisted that RobertKennedy "would not have been unhappy if he [Castro] haddisappeared off the scene by whatever means," and, "I was

just doing my best to do what I thought I was supposed todo."

The murkiness of the record raised a certain problem forthe committee. Either the CIA had undertaken Castro's

murder on its own and was indeed, in Church's words, "arogue elephant rampaging out of control," or Eisenhowerand Kennedy had in fact ordered the CIA to attempt theassassination of foreign leaders, which the associates ofboth Presidents swore they had never done, and wouldnever do. Robert McNamara said he couldn't help thecommittee on this crucial point. He testified that he didn'tremember suggesting Castro's assassination at an SGAmeeting on August 10, 1962, although he did rememberMcCone's telephone call to protest, and he would have totake the committee's word for it that the CIA did, in fact,

try to kill Castro. He didn't know about it.

But McNamara was at the same time meticulous inemphasizing that "the CIA was a highly disciplinedorganization, fully under the control of senior officials ofthe government....I know of no major action taken by theCIA during the time I was in the government that was notproperly authorized .... I just can't understand how it couldhave happened .... " The dilemma was gingerly circledagain and again. Kennedy Administration officials hadnothing but praise for the CIA's discipline; they certainlydid not want to blame the CIA for this: they did not evenwant to blame it on a misunderstanding; and yet they knewthe Kennedys would never have countenanced any suchthing.

The CIA officials involved did not contradict them exactly,but insisted they had the authority, and yet were vaguewhen they tried to explain where the authority came from.More extraordinary still was the restrained way in whichthe high officials of the CIA and of the Eisenhower and

Kennedy administrations treated each other. There was noacrimonious exchange of accusation of the sort one mighthave expected. McNamara, typically, did not want toblame the Agency, and Helms, typically, testified he didn'twant "to take refuge in saying that I was instructed tospecifically murder Castro .... " The claims of both sideswere in soft opposition, and the committee was forced toconfess softly in the end that while it had no evidence thatthe CIA had been a rogue elephant rampaging out ofcontrol, it also had no evidence that Eisenhower orKennedy or anyone speaking in one of their names had


15/49

ordered the CIA to kill Castro. The only indisputable factwas that the CIA did, in fact, try to do so.

The Church Committee reported that it had discovered atleast eight separate plots against Castro of varyingseriousness, ranging from an attempt to give him a

poisoned wet suit for scuba diving to a more determinedeffort, through agents recruited by the Mafia, to poison hisfood.

Some of these plots never survived the first seriousdiscussion, but others were pushed forward over a periodof years, and although none of them came close to success,it was not for lack of effort.

According to Bissell, the first discussion of killing Castrooccurred in the summer of 1960, when planning for the

invasion of Cuba had already been under way for at leastfive months. The early attempts on Castro's life wereassigned to the director of security, Colonel SheffieldEdwards, in August 1960. Edwards and another CIAofficer approached Robert Maheu, a former FBI agent whohad frequently worked for the CIA in the past, and toldhim the CIA would be willing to pay $150,000 for Castro'sassassination. Maheu recommended a Mafia figure namedJohn Rosselli, who agreed to go ahead with the plan, usingother Mafia contacts whose gambling interests in Cuba hadbeen confiscated by Castro in 1959. By October, Rosselli

had recruited Sam Giancana and Santo Traficante, who inturn began to recruit Cubans who might do the job.

The Technical Services Division, meanwhile, was workingon poisons which might be used for the murder, afterGiancana had protested that a gangland-style killing wouldnever work. In a separate but related effort in August 1960,the CIA's Office of Medical Services was given a box ofCastro's favorite cigars and told to treat them with a lethalpoison. They were ready in October, and delivered tosomeone in the Agency -- it is not known to whom -- thefollowing February. The cigars may have been intendedfor Castro during his trip to the United Nations inSeptember 1960. According to David Wise and Thomas B.Ross in The Espionage Establishment, a CIA officer toldMichael J. Murphy of the New York Police Departmentthat the Agency had planned to assassinate Castro with abox of exploding cigars, but then had changed its mind.Perhaps the box of cigars referred to by Murphy was thesame one the office of Medical Serviced had treated withbotulinus toxin by October 7. Perhaps not. In any event,


16/49

the Technical Services Division prepared botulinus toxinpills in February 1961, tested them successfully onmonkeys, and delivered them to Colonel Edwards, whopassed them on to Rosselli in Miami. Late that month orearly in March, Rosselli told the CIA that the pills hadbeen given to a man in Castro's entourage, but that he hadreturned them after he lost his job, and with it his access toCastro. A second attempt in April failed when the agent

got "cold feet," and after the collapse of the Bay of Pigsinvasion, the Maheu-Rosselli operation went into adormant phase.

Early that fall, however, the Mafia plot to kill Castro wasreactivated after Bissell, in a meeting with both Kennedybrothers held in the Cabinet Room, was "chewed out" for"sitting on his ass and not doing anything about getting ridof Castro and the Castro regime." This time Bissellbypassed Colonel Edwards and gave the job to a veteran ofclandestine operations, William Harvey. Earlier in 1961,

Bissell had asked Harvey to organize a unit within theDDP which might recruit agents to carry out assassinationson call -- described with the euphemism "executiveaction," the very phrase, interestingly, which Allen Dulleslater used in his memoirs to describe the "'Murder Inc'branch of the KGB." Harvey organized the group, and onNovember 16, 1961, he and Bissell discussed thepossibility that ZR/RIFLE, the "executive action" group,might be used for killing Castro. Bissell also told Harveyabout the Mafia plot, and later Harvey briefed Helms.

In early April 1962, acting on Helms's explicit orders,Harvey asked Colonel Edwards to put him into contactwith John Rosselli, and a few days later the two men wereintroduced in Miami by the man Edwards had assigned asRosselli's case officer, James O'Connell. Harvey got off onthe wrong foot with Rosselli by telling him to breakcontact on the Castro operation with Robert Maheu andSam Giancana. Harvey had apparently decided the twomen were superfluous and untrustworthy as the result of anepisode eighteen months earlier, in October 1960, when

the CIA-Mafia plot was first getting under way. At thattime, Maheu, as a favor to Giancana, had hired a privatedetective to tap the Las Vegas phone of one of Giancana'sgirlfriends in order to discover if she was being unfaithfulto him. The tap was discovered by a maid, the detectivewas arrested by local police, and Maheu was told to squareit or else.

Later, in April 1961, with the permission of ColonelEdwards, Maheu told the FBI that the tap was connected toan operation he had undertaken for the CIA, and Edwards


17/49

confirmed his story. The problem refused to go away,however, and the following year, in 1962, the Las Vegaswiretap episode helped the FBI to learn the rough outlinesof the plot to kill Castro. This all struck Harvey as aperfect example of an operation going out of control, andhe decided that the first step was to get rid of the clowns,Maheu and Giancana. Rosselli did as Harvey asked, andthe two men met again in New York on April 8, 1962.

Before the end of he month, Harvey delivered four poisonpills to Rosselli in Miami. In May, Rosselli reported thatthe pills were inside Cuba, and later, in June, that a three-man team had been sent in to kill Castro.

But that was as far as things went. By September 1962,when Rosselli told Harvey another three-man team was tobe sent to Cuba, Harvey had concluded that the operationwas going nowhere. He had run the operation with extremesecurity; not even the men who worked for him on TaskForce W (the CIA's end of Operation Mongoose) knew

what he was up to, or where he was going when hedisappeared for a few days every month or two. Bissell hadgiven him the Rosselli operation, Helms told him to give ita shot, Harvey decided on his own it was a will-o'-the-wisp. In February 1963, Harvey told Rosselli the operationwas over.

Harvey's replacement by Desmond FitzGerald and thescuttling of the Rosselli operation did not end but only

redirected the CIA's attempts to kill Castro. One ofFitzGerald's early inspirations was fanciful andimpractical, appealing to his temperamental fondness forthe clever and the ingenious. It called for the TechnicalServices Division to rig an exploding seashell, whichwould be placed on the sea floor in an area where Castroliked to go skin diving. Like many CIA people, in lovewith the subtle and the artful, FitzGerald was fascinated bygadgets and resented skeptics who dourly suggested theywould cost too much or would fail to work or weren't even

needed at all. He was downright petulant at times. WhenSam Halpern once protested that a fancy newcommunications device just wasn't going to work,FitzGerald said, "If you don't like it, you don't have tocome to meetings anymore."

Halpern protested that the seashell plan was inherentlyimpossible to control. How could they be sure that Castrowould be the one to find it? Besides, the bestassassinations do not appear to be assassinations at all,while Castro blowing up on the ocean floor would point a


18/49

finger directly at the United States. Similar protests hadbeen made about the plan to give Castro a box of poisonedcigars. He might hand them all out to a delegation ofvisiting schoolteachers. If the idea was to kill Castro, theyhad to find something which would get him and no oneelse. FitzGerald's ideas weren't turning out any better thanthe earlier ones, such as the proposal to provide Castrowith a poisoned wet suit to be delivered by James B.

Donovan, an American lawyer negotiating the release ofthe Bay of Pigs prisoners. The Technical Services Divisionhad duly purchased a suit and contaminated the breathingapparatus with tubercle bacilli and the suit itself withfungus spores which would cause a chronic skin diseasecalled Madura foot. Critics of this plan claimed that itsauthors had neglected the most elementary points: forexample, the fact that it was in effect a gift from the UnitedStates (the idea was to keep it secret), or Donovan's feelingabout being the gift-giver in this plot. If he didn't know,after all, he might try on the suit himself. As it happened,

Donovan gave Castro a wet suit entirely on his own, andthe CIA's wet suit was destroyed.

But FitzGerald did not abandon the problem. Eventually hecame up with a serious effort to use a major in the Cubanarmy, in contact with the CIA since 1961, named RolandoCubela. Cubela was on intimate terms with Castro, andoften saw and talked to him in his office or at officialfunctions. He and some of his friends bitterly resented theRussian presence in Cuba and felt that Castro had betrayedthe revolution. From the CIA's point of view, he was anideal conspirator, a man with a public reputation as aleader in the fight against Batista, close to Castro,spokesman for a circle of dissidents, and ambitious. On topof that, Cubela had already proved himself as an assassin.In October 1956, he shot and killed the chief of Batista'smilitary intelligence, Blanco Rico, a deed which hauntedhim thereafter and even resulted in a nervous breakdown.Rico had been picked as a target not because he wasruthless or cruel, but because he was a fair, temperate man;he reflected credit on Batista as a leader. Cubela was

convinced that Rico knew why he was being killed, andbelieved that Rico had smiled at him at the very momentCubela pulled the trigger.

The CIA was well aware of Cubela's political and mentalhistory, but decided to use him anyway, since he wasperfectly situated to engineer the one thing which mightactually get rid of Castro -- a palace coup. From thebeginning, Cubela insisted that a coup had to includeCastro's "execution." The word "assassination" disturbedhim; he preferred to say he would "eliminate" Castro. At


19/49

various times he asked the CIA to provide him with exoticassassination devices and more mundane sniper rifles, andthe CIA undertook to give him what he wanted. WithHelms's approval, FitzGerald personally met with Cubelain Paris on October 29, 1963, despite protests fromsubordinates who said that no high CIA official shouldexpose himself in such a manner. Cubela had requested ameeting with Robert Kennedy, but FitzGerald satisfied

him with the claim that he was Kennedy's personalrepresentative. Not quite a month later, on November 22,1963, Cubela's case officer gave him a specially prepared"pen" which might indetectably inject a deadly poison intoCastro; the CIA recommended Blackleaf 40, a widelyavailable toxin which Cubela was to procure on his own.The Church Committee's assassination report says thatCubela dismissed the poison pen as a toy and insisted theCIA could surely come up with something "moresophisticated."

At the end of the meeting, the CIA case officer learned thatKennedy had just been shot in Dallas. During the ensuingtension and uncertainty, the Cubela plot was allowed tolapse for a matter of months. Christmas, 1963, came andwent; nothing happened. Early in 1964, the CIA-Cubelaplotting was revived, and two caches of arms -- one inMarch, the second in June -- were landed in Cuba forCubela's use. That fall Cubela requested a sniper rifle, andthe CIA told him the United States no longer wanted tohave any role in the "first part" of his plan -- that is, inCastro's assassination. Why did the CIA change its mind atthis late date? The record provides no persuasive reason,but it may have been because Lyndon Johnson was quietlysounded out -- so quietly that he may not have known thathe was being asked -- and he wanted no part ofassassination. Clearly, Johnson had not known about theearlier Mafia plots, and Helms did not tell him about theCIA's relationship with Cubela during Johnson's owntenure as President. The important point here is that theCIA's direct involvement in Cubela's assassination planscame to an end at a time when they seemed not to have the

President's sanction.

The subject of assassinations was a painful one for CIApeople. On no other subject did they fight so hard to keepthe secrets, and in particular the secret of presidentialauthority. On this point the testimony of high-level CIAofficials before the Church Committee was elusive in theextreme. Helms in particular remembered next to nothing,and dismissed the rest. He never believed the Mafia plot


20/49

was going anywhere. He let Harvey proceed only to see ifRosselli really had assets in Cuba. Cubela's plan to"eliminate" Castro was indulged to see if he and hisassociates could really put together anything in the natureof an honest plot. The committee had obtained the CIAinspector general's report of 1967, but the memories ofthose involved halted pretty much where the documentscame to an end. The Church Committee's report was

detailed and lawyerly, proceeding point by point in alogical and yet a confusing manner; discussions of closelyrelated events are sometimes scores of pages apart.

But even when one has reassembled the story in its properorder, the picture one gets is fragmentary, occasionallyvivid and complete on minor points, more often bald andout of focus. The primary reason for this is the tendency ofCIA officials to suffer memory lapses on all those points,which were very numerous, that had not survived in thefiles. In addition, of course, Eisenhower, both Kennedys,

both Dulleses, General Cabell, and other high officials haddied. Livingston Merchant and Admiral Arleigh Burkewere too ill to testify. Some of the lower-level officials --William Harvey, Justin O'Donnell, Sidney Gottlieb, andothers -- testified at length but did not really know whogave the orders or when, and would not have presumed toask.

The idea of assassination itself did not seem to trouble CIAofficials who testified. The wisdom of the undertaking wassomething else again. It was stupid, foolish, ridiculous,unworkable; worse than a crime, a blunder -- the regularspiel. Everyone had his own adjective, none of themflattering. The best they could muster by way of

justification was "the climate of the time," the Kennedys'hysteria on the subject of Castro, the eager willingness ofthe Cubans who were recruited seriatim to do the job. Butall the same, they shook their heads in dismay. More thananything else, it seemed to be the sheer difficulty ofassassination -- that is, of a genuinely secret assassination -- that left them wondering.

But on the question of presidential authority, there is nosuch equanimity. One exception said that no one in theCIA doubted for a minute that Eisenhower and Kennedy"jolly well knew," but others, more closely involved, didmore than simply squirm in their chairs. Several differentmen, in fact, showed dramatic signs of psychological stressin discussing this point.


21/49

It is inconceivable that Richard Helms would ever betrayhimself in so unmistakable a manner. But in his testimonybefore the Church Committee, Helms more than oncerevealed an uncharacteristic degree of irritation with thecommittee's insistent return to the question of authority.He was being as clear as he could: the Kennedys wantedCastro out of there, the CIA did not go off on its own inthese matters, the Agency was only trying to do its job.

What more could he say? Senator, how can you be sogoddamned dumb? This isn't the kind of thing you put inwriting.

And despite the Church Committee's diligent search, theynever did find anything in writing. The committee didlearn, however, of three separate occasions when one orboth Kennedys discussed the assassination of Castro in amanner indicating that it lay heavily on their minds. Thefirst occasion occurred in March or early April of 1961,

just before the Bay of Pigs invasion, at the height of the

first Rosselli effort to poison Castro, when PresidentKennedy asked his friend Senator George Smathers whathe thought the Latin American reaction would be to theassassination of Fidel Castro. Smathers said he toldKennedy the murder would be blamed on the United Statesand on Kennedy personally, and that he, Smathers, wastherefore against it. According to Smathers, Kennedyimmediately responded that he was against it too. But inmid-March 1961, before their conversation, the CIA hadalready given botulinus toxin pills to Rosselli in Miami,and a second batch were to be handed over on April 21.

The elimination of Castro was raised again by theKennedys -- in more ambiguous terms this time -- during ameeting with Bissell in September 1961. Bissell laterdescribed the meeting to his Cuban desk officer in mid-October. He said he had been called to the White Houseand "raked stem to stern" by both Kennedys in the CabinetRoom, and by Robert Kennedy in particular. By allaccounts, both Kennedys could make a point when theywanted to. McGeorge Bundy, who told the Church

Committee that ordering an assassination would have been"contrary to everything I know about their character," alsosaid that when there "was something that they reallywanted done they did not leave people in doubt." LarryHouston, who had on May 7, 1962, briefed RobertKennedy about the early, pre-Bay of Pigs Mafia plot, madethe same point: "If you have seen Mr. Kennedy's eyes getsteely and his jaw set and his voice get low and preciseyou get a definite feeling of unhappiness." The Cuba deskofficer got a clear impression from Bissell's description ofwhat he'd been told by the Kennedys: they wanted the CIA


22/49

to get rid of Castro, and they meant get rid of Castro.

Castro continued to be on the Kennedys' mind that fall of1961. On November 9, the attorney general took TadSzulc, then a reporter for the New York Times, to meet thePresident, who asked Szulc, "What would you think if Iordered Castro to be assassinated?" Szulc told thePresident it wouldn't work, and that the United States

should not do such things. Kennedy said he and his brotherfelt the same way. In Szulc's note of the conversation,made the same day, he wrote "JFK said he raised questionbecause he was under terrific pressure from advisers (thinkhe said intelligence people, but not positive) to okay aCastro murder. Said he was resisting pressures."

Despite all the evidence gathered by the ChurchCommittee, it never found anything like an order to killCastro in writing, and it never found a witness who wouldconfess explicitly that he had received such an order. The

committee's response to the incomplete record was to leavethe question of authority hanging. Must we do the same?Lacking a smoking gun in the form of an incriminatorydocument or personal testimony, we can reach no firmconclusion, but at the same time the available evidenceleans heavily toward a finding that the Kennedys did, infact, authorize the CIA to make an attempt on Castro's life.

The evidence is particularly persuasive on two points.First, President Kennedy's conversations with SenatorSmathers and Tad Szulc on the subject of assassination

both occurred at times when the CIA was actively trying tokill Castro with the aid of the Mafia. Second, the briefingof Robert Kennedy by Lawrence Houston and SheffieldEdwards elicited a very narrow response from Kennedy.The facts surrounding the briefing, held on May 7, 1962,are extremely complex, but at its heart the episode is asimple one: a case of the dog that didn't bark. Houston toldthe committee that Kennedy's anger was directed at theCIA's use of the Mafia; He made the same point even moreemphatically to me. "Kennedy was mad," he said. "He was

mad as hell. But what he objected to was the possibility itwould impede prosecution against Giancana and Rosselli.He was not angry about the assassination plot, but aboutour involvement with the Mafia." Perhaps Kennedy did notknow the whole story, Houston conceded, but he added:"All I know is that [Robert] Kennedy knew about one ofthem [the assassination plots] in very great detail."

The record is clear, then, that Kennedy was thoroughlybriefed about the details of an attempt to murder Castroduring his brother's presidency. The record is clear that the


23/49

attempt to kill Castro continued. And the record is clearthat, despite his knowledge of the earlier attempt, RobertKennedy did not protest to the CIA, to its director, JohnMcCone, to Helms, or to anyone else in the Agency aboutthat attempt. He was mad about the use of the Mafia.Period. Would he have kept his mouth shut, and donenothing, if he had discovered that the CIA, answerable tohis brother, had tried to murder a foreign leader without his

brother's approval? It seems unlikely.

That was the first time the dog didn't bark. The secondtime occurred during the Church Committee'sinvestigation, when Kennedy Administration officialsmight have been expected to be publicly furious at theCIA -- an executive agency, as Helms often remindedCongress when he was director -- for undertaking anythingso fundamental as an assassination without the President'sexplicit approval. Instead, they said that the Kennedys theyknew would never have done such a thing, and left it at

that. Why were they so complaisant? Well, you can push aman keeping a secret just so far.

4. A Case History: Chile

DURING the 1960s, Chile received more American aidper capita than just about any other country in the world --Vietnam excepted -- and the CIA provided half the money

spent in the 1964 election won by the ChristianDemocratic party candidate, Eduardo Frei.

Frei was the beneficiary, not only of CIA funds givendirectly to his party (something he did not know), but of aCIA propaganda program intended to scare the livingdaylights out of Chileans at the prospect of a victory bySalvador Allende, whose Popular Action Front wasdepicted as nakedly Stalinist. Posters of Russian tanks inthe streets of Budapest and of Cubans in front of Castro'sfiring squads proliferated on Chilean walls in 1964. CIA

assets in the Chilean press hammered on the same themewhile CIA election experts coached Christian Democraticparty workers on American media and get-out-the-votetechniques. A quieter but equally effective CIAdisinformation effort helped to divide the left and to keepAllende defending himself against charges which werefalse or half true or even all true -- such as foreign fundingof his party -- but which were equally true of his principalopponent, Frei.

In the end Frei's victory in 1964 was probably his own, but


24/49

not its margin: the credit for that must go to the air of crisisthat polarized the Chilean electorate, and which had beenlargely the CIA's doing.

As early as April 15, 1969, Helms warned Henry Kissingerthat an early start was necessary if the CIA was to repeat in1970 its successful role in the 1964 election. Kissingerdecided to let the matter ride for the moment. The situation

was complicated by a dispute between the CIA and U.S.Ambassador to Chile Edward M. Korry, who favoredminimal interference. In Santiago, CIA station chief HenryHeckscher wanted to support the rightist candidate, JorgeAlessandri, directly. Korry balked at that, but Heckscherpersuaded him that his hands-off policy was suggestingAmerican indifference to the cause of democracy, and wasin effect helping Allende, who was receiving Russianfunds. Korry and Heckscher then drafted a joint plan for ageneral anti-Allende campaign which would continue tobar direct support for any single candidate. At length a

proposal for anti-Allende "spoiling operations" was finallyapproved on March 25, 1970.

At this point the multinational companies intervened. Theywanted not a general anti-Allende, scare-the-peoplecampaign, but a more aggressive program of positivefinancial and technical support for Alessandri, the onlycandidate in the election who opposed expropriation. OnApril 10, a group from the Business Council on LatinAmerica met with the assistant secretary of state for inter-American affairs, Charles Meyer, to urge a major pro-Alessandri effort. Meyer, a former Sears, Roebuckexecutive in Latin America who had been given his StateDepartment job through the influence of DavidRockefeller, was studiously noncommittal when thechairman of the board of Anaconda, C. Jay Parkinson, saidhis and other interested American companies were willingto put up $500,000 to block Allende. Another StateDepartment officer present at the meeting, WilliamStedman, sent Korry a memorandum describingParkinson's offer, and Korry responded on April 28 with a

stinging cable arguing against any such involvement byU.S. business, claiming that Alessandri was a candidate ofthe rich, who could well afford to pay for their champion'scampaign, and repeating again that U.S. support for arightist was going to backfire against the United States.

Deflected for the moment by Korry's opposition, themultinationals changed their strategy. Instead ofproceeding through the State Department, they decided toenlist the aid and expertise of the CIA. In May 1970, JohnMcCone, who had appointed Helms head of the DDP back


25/49

in 1962, and who was now a member of the board ofdirectors of ITT, approached Helms privately to discuss aCIA-ITT program to support Alessandri. As DCI back in1964, McCone had refused an ITT offer of funds for theChilean election, but now he was ready to propose whatamounted to the same thing. In July 1970, McCone againcontacted Helms, who in turn arranged a Washingtonmeeting between William Broe, chief of the Western

Hemisphere Division of the DDP, and Harold Geneen, thehead of ITT. Geneen offered Broe and the CIA $1 millionin ITT funds for a pro-Alessandri campaign.

The various congressional committees that investigated theChilean episode cited the ITT offer but never fullyexplained what it was for. Giving the CIA money, after all,is bringing coals to Newcastle. Why was the offer made?Both Korry and the State Department opposed a pro-Alessandri campaign, and Kissinger duly limited the U.S.effort to an anti-Allende campaign. Rejecting State

Department reservations at a 40 Committee (the WhiteHouse CIA Oversight Committee) meeting on June 27,1970 -- "I don't see why we have to let a country goMarxist just because its people are irresponsible" --Kissinger nevertheless restricted the U.S. effort to spoilingoperations and a $500,000 contingency fund, proposed byKorry ten days earlier, to "influence" the final vote of theChilean Congress, should Allende win the election onSeptember 4. The weakness of Kissinger's strategy, asHelms suggested without emphasis at several meetings on

Chile, was the weakness of any political campaign whichproposed to beat somebody with nobody. It is hard not toconclude then, that ITT's million-dollar offer, madeindirectly through McCone, was actually an attempt toreach a working agreement with the CIA for a pro-Alessandri campaign which was to remain secret fromKorry and the State Department, and perhaps -- but this isless likely -- even from the White House itself.

Did McCone, a former CIA director, have reason tobelieve that the CIA would lend itself to any such free-

lancing scheme? In any event, the CIA cooperated in amodified version of such a scheme, providing ITT with thenames of Chileans through whom ITT could supportAlessandri on its own. According to several sources, theCIA went further, and provided ITT with localintroductions as well. Thus the CIA was, in effect,operationally supporting a policy which had beenspecifically rejected, so far as the record shows, by theU.S. government. The bald facts of this arrangement werecited by the Church Committee but then more or lessignored, an omission which was to infuriate Korry later.


26/49

In the early Summer of 1970, the CIA, over Korry'sprotest, managed to persuade the State Department tosupport a pre-election poll in Chile. The result was a CIAprediction that Alessandri, would win with 42 percent ofthe vote. Korry took issue and reviewed the poll with thehelp of embassy officers. They cabled the StateDepartment criticizing the CIA for basing its poll on the1960 Chilean census, and concluding that Alessandri

would win 40 percent of the vote at best, and likely a gooddeal less. The CIA reviewed the review and stuck to itsoriginal figures; Alessandri would win with 42 percent.

But Alessandri did not win on September 4, despite a CIApropaganda effort which was a replay of the 1964 scarecampaign. The actual results were Allende 36.3 percent,Alessandri 34.9, and Rodomiro Tomis 27.8. The reactionon the right in Chile, among the multinationals, and in theWhite House was all but identical: alarm verging on panic.Nixon and Kissinger, perhaps lulled by the CIA's poll into

a relative low-key intervention, now felt betrayed anddesperate: something had to be done to stop Allende. Thissentiment was fully shared by the multinationals. TheChilean publisher Agustin Edwards, a longtime ally of theCIA, asked Henry Heckscher to arrange a meeting withKorry at the embassy. There Edward bluntly asked, "Willthe U.S. do anything militarily -- directly or indirectly?"Korry was as unhappy about Allende's victory as Edwards,but he was dead set against anything in the nature of acoup to keep Allende out of office. He told Edwards that

the United States intended to abide by the election results.But Edwards had other avenues to the U.S. government,and he immediately used them. As the owner of a localPepsi-Cola bottling plant, he knew PepsiCo's chief, DonaldKendall, an old ally and friend of Nixon. Edwards fledChile, met with Kendall in the United States, andprophesied general disaster if Allende was allowed to takeoffice. Kendall was impressed and arranged for HenryKissinger and John Mitchell to meet Edwards at a privatebreakfast on the morning of September 15. 1970. A weekearlier, Harold Geneen of ITT, also alarmed, had asked

McCone to get in touch with Helms again, but this timeHelms delayed his response, waiting to see what the WhiteHouse wanted to do.

Korry, meanwhile, had picked up wind of a possiblemilitary coup as a means of preventing Allende'sconfirmation by the Chilean Congress in its vote scheduledfor October 24. The commander of the Chilean armedforces, General Rene Schneider, was known to be firmlyopposed to any unconstitutional attempt to block Allende'sconfirmation. Since the birth of Chilean independence in


27/49

1818, democracy in Chile had been interrupted on onlythree brief occasions, the last in 1932 -- a remarkablehistory in Latin America, and one which Schneider wantedto maintain.

Other military officers, however, were not so punctilious.For nearly eight years, the CIA had been painting a leftistvictory in the darkest possible light, and elements of the

Chilean military, like Chilean businessmen hurriedlyexporting their capital abroad, were afraid that Stalinismwas around the corner. One of the early militaryconspirators was Brigadier General Roberto Viaux, wholeft the army after an abortive coup called the "Tacnazo" in1969, and who was an erratic, politically irresponsible manwith a beautiful, ambitious wife. Korry had ordered theCIA to refrain from all contact with Viaux and othermilitary conspirators, and he later barred two local ITTmen from the embassy, Hal Hendrix and Robert Berrellez,because of their plotting with the Chilean right. On

September 12, responding to a Kissinger-40 Committeerequest for a "cold-blooded assessment" of the situation,Korry cabled that "our own military people [are]unanimous in rejecting possibility of meaningful militaryintervention .... What we are saying in this 'cold-bloodedassessment' is that opportunities for further significantUSG action with the Chilean military are non-existent."

Two days later, on September 14, the 40 Committeedecided to risk what the CIA referred to as "the RubeGoldberg gambit," an unwieldy scheme to (a) persuadeFrei to resign, (b) have his vice president succeed to thepresidency, and then (c) "influence" -- with a $250,000CIA contingency fund -- the Chilean Congress to vote forFrei, who was otherwise constitutionally ineligible tosucceed himself. Korry went along with this improbablescheme on the grounds that it depended on Frei, and thusoffered a "Chilean solution." But Heckscher had alreadywarned Korry that nothing of the sort could work, sinceCIA agents had learned that Tomic and Allende hadreached a secret deal to back the leader if either of the two

candidates should place first or second in the election. Ineffect, they were collaborating to beat the right. Such adeal could hardly have been reached without Frei's supportas leader of the Christian Democrats, but Korry refused tobelieve the CIA was right in its report of the deal, andwhen he learned later that the Agency had been right, hefelt something of a fool for ever having approached Freiwith the Rube Goldberg gambit.

In any event, the gambit went nowhere, and while Korrycontinued to urge Frei to think of something, at the same


28/49

time he peppered Washington with warnings that only Freiand a "Chilean solution" had any chance at all, and that amilitary coup by the likes of Viaux would be the height offolly. But coup rumors continued to circulate, and Korry'ssuspicions were aroused. One day Heckscher -- "thisnormally courteous man," in Korry's words -- suddenlyblew up in anger at Korry's low-key intervention with Frei,an explosion the more remarkable because the two men

were not alone, but accompanied by Korry's deputy chiefof mission, Harry Shlaudeman. "Why the hell don't youtwist Frei's arm?" Heckscher shouted. "You're tellingWashington you're doing it and you're not!" Korry warnedHeckscher that he'd be out in twenty-four hours if he didnot calm down, and then lectured him that it was up to Freiand the Chileans to block Allende. If they couldn't find away, the United States couldn't do it for them.

Heckscher later apologized, but Korry began to wonder ifthe CIA wasn't up to something behind his back. He asked

Shlaudeman to look into it, and Shlaudeman reported thathe could find no evidence that the CIA was plotting withthe military on its own. He told Korry he was beingparanoid. Korry was not.

The plot called Track II had begun with a meeting in theOval Office of the President, Kissinger, John Mitchell, andHelms on September 15, 1970, just one day after Korryhad been ordered to pursue the Rube Goldberg gambit with

Frei in Santiago. Helms testified later that he thoughtNixon's determination to act was the doing of DonaldKendall and Agustin Edwards, who had met Kissinger forbreakfast that morning. Helms knew Kendall fairly well,having seen him at Washington meetings perhaps four orfive times a year, and he knew that Kendall and Nixonwere close, Kendall having given Nixon his first bigcorporate account after Nixon began practicing law in NewYork. But more immediately, Helms had been asked byNixon or Kissinger -- he can no longer remember which it

was -- to meet with Kendall and Edwards at a Washingtonhotel. The two men made quite an impassioned appeal forCIA help in blocking Allende, and Helms concluded thatthey must have made the same appeal to Nixon, with somesuccess.

Nixon himself cited a different source for his concernabout Allende. He told David Frost that it began with aconversation with an Italian businessman who warned him,"If Allende should win the election in Chile, and then youhave Castro in Cuba, what you will in effect have in Latin


29/49

America is a red sandwich and eventually it will all bered." Whatever the exact source of Nixon's fears, at thatSeptember 15 meeting in the Oval Office he made nosecret of his determination to stop Allende. He outlined thedangers as he saw them, swore his Administration wouldnot "cave in at the edges," and told Helms to leave nostone unturned in the attempt to block Allende'sconfirmation. "If I ever carried a marshal's baton in my

knapsack out of the Oval Office," Helms told the ChurchCommittee "it was that day." He also carried a single pageof handwritten notes which capture the tone of hisinstructions:

One in 10 chance perhaps, but save Chile!worth spendingnot concerned risks involvedno involvement of embassy$10,000,000 available, more if necessaryfull-time job -- best men we have

game planmake the economy scream48 hours for plan of action

At that point Helms thought a one-in-ten chance forsuccess was optimistic, and nothing happened later toimprove the odds. Thomas Karamessines, deputy directorof plans, felt the same way, and so did David Phillips,brought back from Brazil to head a special Chile TaskForce for the duration of the operation. Henry Heckscherwas even more pessimistic, and he peppered Langley withhis doubts to such a degree that on October 7 he wasordered to stop protesting and limit his cables to what hedid. When Heckscher continued to balk, Karamessinesordered his return to Washington. "Well," Heckscher told afriend at Langley, "I guess I've lost my job." He was notfired, but he was most unmistakably "read the riot act,"according to several sources. This was something the CIAhad been told to do, Langley was committed to giving it atry, and Heckscher was expected to bite the bullet.

Track II went forward, then, despite the unanimouspessimism of those most closely involved, because Helmshad his marching orders from Nixon and Kissinger."Nobody," said Karamessines, "was going to go into theOval Office, bang his fist on the table, and say, We won'tdo it." The only limits Helms imposed on the operationwere those demanded by security: he was willing enoughto try and fail, not at all ready for the failure to becomepublic. Despite Korry's fear that an attempted coup mightbecome another Bay of Pigs, not a word surfaced fornearly five years, and the operation emerged then only


30/49

because another branch of the government discovered anoutline of the facts and insisted on publishing them.

Secret or not, failure is failure, and Heckscher had noenthusiasm for a project with so little chance of success.The trouble, in his view, was that the CIA had nothing to

work with. The local station was heavily dependent on theembassy's defense attache, Colonel Paul Wimert, for itscontacts with Generals Roberto Viaux and CamiloValenzuela and their co-conspirators, largely becauseKorry had forbidden the CIA to keep in touch withdissident military officers. Not only was the CIA forced todeal indirectly with the conspirators, at least in thebeginning, but as assets they weren't exactly formidable.Viaux was an unreliable ally. Heckscher knew, andreported, that Viaux's circle had been infiltrated by theChilean MIR, an organization of the extreme left, and as

time went by it grew increasingly apparent that neitherViaux nor Valenzuela had a plausible plan for takingpower. The best they could come up with was a successionof jerry-built schemes to kidnap General Schneider in thehope that Frei, or the rest of the Chilean militaryestablishment, might decide to act in the ensuing crisis.But even that scheme percolated erratically, despite a CIAoffer to pay $50,000 for Schneider's successful abduction.

Helms and Karamessines informed Kissinger and his aide,Alexander Haig, of the bleak picture on a regular basis. Atthe end of September, Helms sent William Broe to askEdward Gerrity of ITT for help in making the Chileaneconomy "scream," but now ITT had cold feet and refused.Not long after, Viaux had to be dissuaded from apremature coup attempt which might wreck everything.Kissinger later told the Church Committee that a gloomyKaramessines report on October 15 led him to cancel thewhole operation. Karamessines did not remember it quitethe same way. The Viaux approach was abandoned, hesaid, but at the same time Kissinger ordered the CIA to

keep the pressure on "every Allende weak spot in sight --now ... and into the future until such time as new marchingorders are given." Two days later, according to one source,Karamessines was called in by Nixon and told to find amilitary alternative to the hopeless Viaux.

On the same day, October 17, in Santiago, a CIA officertold Viaux not to push too fast, while Colonel Wimert metwith another group of Chilean military conspirators whoasked him for eight to ten tear-gas grenades, three .45-caliber submachine guns, and 500 rounds of ammunition,


31/49

claiming they were needed for self-protection. Wimertobtained the grenades from the CIA and delivered them toan associate of Valenzuela, who, later the same day, toldWimert that coup plans were now ready, and would beginthe next night with the kidnapping of Schneider followinga military dinner. The plan came to nothing whenSchneider left the dinner in a private car, well guarded bypolice. Wimert was told another attempt would be made

the following night, on October 20, but that too failed, andHeckscher concluded that time had run out.

Nevertheless, Wimert delivered the promised machineguns to Valenzuela's associate at a 2 A.M. meeting onOctober 22. Five hours later, a group of militaryconspirators met for final planning of a last attempt toabduct Schneider, and at 8 A.M. they halted the general'scar. Schneider attempted to resist, drew his revolver, andwas shot and fatally wounded by his would-be abductors.He died three days later, one day after Allende's

confirmation.

The Church Committee's description of Schneider's murderwas punctiliously factual. Because Schneider was killedwith handguns and because the military officers to whomColonel Wimert gave the machine guns were not present atthe 7 A.M. meeting before the botched kidnapping attempt,and because it was General Viaux who was heldprincipally responsible for the fatal attempt by the Chileancourts later, the committee concluded that the CIA was notimplicated directly in Schneider's death. The trouble withthis highly legalistic arrangement of the facts is that itobscures three points: (1) there was no clear line ofdivision between the Viaux and Valenzuela circles, and theChilean courts also held the latter responsible, though to alesser degree; (2) the failed attempts of October 19 andOctober 20 had been carried out by the same group whichfatally wounded Schneider on October 22, althoughWimert's discussion of those attempts had been withValenzuela, a fact which suggests the two generals were

acting in close concert; (3) both Viaux and Valenzuelawere in regular contact with the CIA, were activelyencouraged to proceed with their plan for kidnappingSchneider, were promised a substantial sum of money ifsuccessful, and very likely would have done nothing at allwithout American encouragement to move. If the CIA didnot actually shoot General Schneider, it is probably fair tosay that he would not have been shot without the CIA.

The day before the Chilean Congress was to vote toconfirm the next president, Helms, back from a trip to


32/49

Vietnam, met in Langley with the Chile Task Force for adiscussion which was mildly hopeful that the plan mightstill work. "It was agreed," a CIA memorandum of themeeting said, "that a maximum effort has been achieved,and that now only the Chileans themselves can manage asuccessful coup. The Chileans have been guided to a pointwhere a military solution is at least open to them."

But it didn't work out that way, just as Heckscher hadpredicted in a cable to Langley as early as October 9. TheChilean military rallied behind General Gonzalez Prats,Schneider's successor, and despite the fact that GeneralValenzuela was appointed commander of Santiagoprovince, there was no coup. Allende was confirmed onOctober 24.

Nixon and Kissinger were not happy with the events of

September and October 1970. Far from being grateful toHelms for having made such a determined effort withoutso much as a word leaking to the press, they blamed himfor Allende's victory. Kissinger personally asked thePresident's Foreign Intelligence Advisory Board (PFIAB)to make a special investigation of the Chilean episode, andat the same time word began to spread around town thatthe Administration was unhappy with the Agency. InDecember 1970, John McCone paid one of his regularvisits to Langley and dropped in to see John Bross, whohad handled the CIA's explanations to the PFIAB. McConesaid he'd been to see Kissinger. "Everybody's very downon Helms for failing to take drastic action to stop Allende,McCone said.

Bross asked Helms about this and Helms confirmed thatthe Administration was indeed unhappy, thinking he'dfailed to warn them in time of the likelihood of an Allendevictory, and then had failed again to block Allende'sconfirmation after the election. But in Helms's view thefailure belonged at least equally to the Administration, for

paying no attention when he warned the 40 Committee atleast a year ahead of the election that then was the time forthe CIA to get involved, and to Ed Korry, for resisting apro-Alessandri campaign down to the bitter end.

I never got up and pounded the table and saidyou've got to take drastic action," Helmsconceded to Bross. "I don't think that was myrole. That's what we're always being criticizedfor -- intervening in policy.


33/49

Helms thought it unfair that he should be singled out forblame in the Chilean fiasco, but at the same time heconsidered the matter as akin to an argument in the family,and whenever it threatened to go outside of the family, hedid what he could to keep the whole episode secret. He hadthrown Fulbright off the scent back in September 1970,when he evaded Fulbright's question about CIAinvolvement with Chile, and simply remarked that if the

CIA had really put its weight into the election, thingsmight have turned out differently -- which was very likelytrue, but not, as the lawyers say, responsive. He hadsidestepped an invitation to testify from Senator CharlesPercy on February 5, 1973. He completely misled SenatorSymington two days later. When he was called back fromIran, where be was U.S. ambassador, to testify on May 21,1973, he narrowly escaped a list of 100 questions preparedby the Foreign Relations Committee staff when the hearingwas held in public, a maneuver which guaranteed that thesenators -- not the staff who had been studying the Chilean

episode -- would be asking the questions.

It was not until January 1975 that Helms was finallycornered and forced to explain his earlier evasions. Helmsexplained that the CIA hadn't given money directly toAllende's opponents, that the CIA didn't try to fix the votein the Chilean Congress because investigation had shownit couldn't be arranged, that the CIA didn't try to overthrowthe Chilean government because the Agency failed to findanyone who could really do it. If there are explanations

which can be called lame, these are cripples. Helms hadgiven Symington the same "explanations" the night beforehis testimony back in May 1973, and Symington, a friend,had been content with them. But others preferred todescribe Helms's testimony by a balder term -- lies.Enough people subscribed to this definition to move thewhole question to the Justice Department, but the heart ofHelms's explanation was more to the point.

AMBASSADOR HELMS: I realize, sir ... thatmy answer [to a question about the attempt tobribe the Chilean Congress] was narrow, but Iwould like to say something here. I didn'tcome into the Multinational Committee[headed by Frank Church, where Helmstestified on Chile on March 6, 1973, a fewdays before leaving for Iran] hearing tomislead you, but I have had as Director ... a lotof problems, and one of the principalproblems was who in the Congress [I] was


34/49

really to divulge all of the details of covertoperations to, and I must say this has given mea great deal of difficulty over the years .... If Iwas less than forthcoming it wasn't because Iwas being bloody-minded, it was simplybecause I was trying to stay within what Ithought was the congressional guidelines.

That was as close as Helms ever came to saying that hisinterrogators had no right to honest answers, because theyhad no right to ask the questions they had. But by that timeit was not primarily the senators whom Helms had tosatisfy. The nature of his testimony -- narrow in the line ofduty? so evasive as to pass into the realm of lies? -- was nolonger academic. The matter had been referred to theDepartment of Justice, and the man who had hand-delivered the documents in the case was someone Helmshad helped to rise in the Agency, someone who might havebeen considered to a degree in Helms's personal debt for

his position as director of central intelligence, WilliamColby. Helms's fight was not really with the senators bythis time -- with the possible exception of Church, they hadlittle appetite for Helms's blood -- but with Colby's policyof letting out the "bad secrets." The very first result ofexposing the "bad secrets" -- others, of course, werecoming -- was a charge of perjury leveled against RichardHelms.

5. Family Jewels

THE men who followed Helms did not share his regardfor secrets, for their inviolability. The combination ofWatergate and James Schlesinger would crack open theAgency's secret past, and William Colby would finish the

job. Schlesinger arrived in December 1972 with a mixtureof suspicion and contempt for the "gentlemen's club" thathad wielded power in the CIA since the 1940s. Schlesinger

had a great many ideas, but at their heart was a plan to gutthe clandestine services. "That DDP, that's Helms'sPraetorian Guard," Schlesinger told the London chief ofstation, Rolfe Kingsley, during a trip to England. "I'mgoing to bust it up."

One of Schlesinger's first acts as director was to hold ameeting of DDP people in the Agency's main auditorium.From now on, he said, intelligence is going to be a twenty-year career. It's time to give way to young blood.Schlesinger was going to clear the place out. The process


35/49

was brutal, but even many CIA people concede it was longoverdue.

Like Helms, Schlesinger held a regular morning meetingwith his deputy directors in his office, and every morninghe wanted to see numbers. He didn't want excuses; hewanted the names of the people who were going. EdProctor, the deputy director for intelligence, was frequently

criticized for moving too slowly. Carl Duckett, the deputydirector for science and technology, did better; he came inwith a list of names every day and the men under himbegan openly calling him a heartless son of a bitch.William Colby, who had been appointed to replaceThomas Karamessines as head of the Deputy Directorateof Operations (formerly the DDP), came back to his officeregularly with an echo of Schlesinger's complaint: "Wearen't getting any numbers." He gave the job to GordonMason, chief of the DDO's Career Management Group,apparently hoping to insulate himself from the harsh

decisions Schlesinger demanded. But Mason refused to letColby off the hook. He picked his candidates for the axcarefully, but once he had put together a pile of personnel

jackets, he brought them to Colby and said, "Here they are,you make the decisions."

Schlesinger did not remain long at the CIA; On May 9,1973, Nixon appointed him to replace Elliot Richardson atthe Department of Defense, who was replacing RichardKleindienst at the Department of Justice, who wasresigning because his old friend John Mitchell was finallyfacing indictment for his role in the Watergate scandal. Butduring Schlesinger's brief tenure as DCI, the shortest in theAgency's history, he fired more than a thousand officersthroughout the Agency, more than a hundred of them oldsoldiers in the DDP/DDO.

The firings came in waves: If the pace wasn't briskenough, he would do the job himself, going down a list ofofficers and saying, "He's been here twenty years that'slong enough, out." It was a crude method, and it got rid of

some able CIA officers along with the dead wood, butSchlesinger could not be argued with.

Yet if Schlesinger was resented as an outsider, WilliamColby came to be disliked by many (not all) CIA people assomething even worse, a kind of traitor who betrayed thetrust Helms had shown in him, and who severely damagedthe Agency during the two and a half years he ran it.Helms had given Colby just about every important job he'dheld. But Colby was fundament

Documents

Inside the Department of Dirty Tricks