Book
Legacy of Ashes: The History of the CIA
Weiner, a New York Times reporter, narrates the CIA from its creation in 1947 to 2007 in six parts keyed to presidents and directors, from Truman to George W. Bush. He draws on declassified CIA, State Department and National Security Council records, oral histories and interviews, with notes after each chapter. The book argues that the agency never built a first-rate espionage and analysis service, that presidents ordered covert action to stand in for knowledge, and that directors misled presidents and Congress to protect the agency. Chapters run as episodes: Iran 1953, Guatemala, Cuba, Vietnam, the Church-era disclosures, Iran-Contra, the Iraq estimate, September 11 and the post-2001 detention program.
Who made it
- Author is a reporter for The New York Times who won a Pulitzer Prize; CIA historian Nicholas Dujmovic's review describes him as 'a Pulitzer Prize-winning journalist for the New York Times'.
- Doubleday is an imprint of Knopf Doubleday Publishing Group (Open Library publisher listing). The book won the 2007 National Book Award for Nonfiction (National Book Foundation).
- Weiner appears in his own narrative as a reporter: he quotes his New York Times article of 7 October 2001 and describes being in Jalalabad in November 2001.
- Dujmovic's review was published in Studies in Intelligence, vol. 51 no. 3 (2007), the CIA Center for the Study of Intelligence journal. Dujmovic writes 'As a CIA historian'. The review carries the journal's disclaimer that the author's views are not an official position of any component of the U.S. government.
- Weiner's acknowledgments give 'a tip of the hat to the men and women of the history staff' of the CIA (as quoted in Dujmovic, note 16). Most of the narrative rests on interviews, oral histories and declassified CIA internal histories.
Its claims checked against the record
14 claims checked:6 supported · 6 partly supported · 2 contradicted · 0 not in the record
A verdict rests on the documents named in its row.
| The work says | Verdict | What the record shows |
|---|---|---|
| Eisenhower's phrase 'legacy of ashes' was his anger at Allen Dulles and the CIA: at the 'final gatherings' of his National Security Council, after Dulles called the intelligence board's proposals 'preposterous', 'madness' and 'illegal', the president 'exploded' and said he would 'leave a legacy of ashes' to his successor. The book takes its title from this scene. | Contradicted by the record | The 5 January minutes read: 'He had suffered an eight-year defeat on this question but would leave a legacy of ashes for his successor.' The question was the separate Army, Navy and Air Force intelligence organizations, and the minutes record the president saying 'it was the function of CIA to acquire strategic intelligence'. The minutes do not record him addressing Dulles or criticizing the CIA. |
| The CIA and British intelligence organized the August 1953 overthrow of Iran's prime minister Mossadeq. The CIA paid street crowds, politicians and officers, and the shah returned to power afterward. | Supported by the record | The briefing describes ruffians paid to demonstrate on 19 August, anti-Mossadeq forces assaulting his house with tanks, and Zahedi speaking from the radio station. Weiner's account of the planning (Operation Boot/Ajax, Roosevelt in the field) matches the Wilber history the briefing reproduces. The casualty and cash figures in the book (for example 'at least a hundred people died') are not in the briefing. |
| The CIA's 1954 operation against Guatemala's president Arbenz (PBSUCCESS) nearly failed, and the agency improvised air attacks and propaganda to save it. | Supported by the record | The briefing reports a $2.7 million PBSUCCESS budget, CIA lists of officials to 'eliminate immediately in event of [a] successful anti-Communist coup', and Cullather's account that 'Operation Success' barely succeeded while the CIA presented it as an unblemished triumph. |
| The CIA task force judged by November 1960 that its Cuba invasion concept could not succeed without U.S. forces. Richard Bissell did not tell President-elect Kennedy, and Eisenhower had never approved an invasion. | Supported by the record | The briefing states that a 15 November 1960 task force meeting judged the original concept unachievable and that this was not shared with Kennedy then or after inauguration. It quotes Kirkpatrick's finding 'a failure to advise the President that success had become dubious' and 'plausible denial was a pathetic illusion'. |
| The Gulf of Tonkin attack of 4 August 1964 did not happen, and National Security Agency analysts skewed the signals intelligence to fit the claim of an attack. Congress passed the Tonkin Gulf Resolution on that basis. | Supported by the record | The National Security Archive release of the Hanyok study says NSA provided top officials with 'skewed' intelligence supporting claims of an August 4 attack, describes 'an active effort to make SIGINT fit the claim', and states there was no second attack on 4 August. The Hanyok PDF is image-only here, so Weiner's quoted phrases ('deliberately skewed', 'firing over 300 rounds madly') and the account of the destroyed intercept are not checked word for word. |
| In 1967 the CIA accepted a falsified figure of 299,000 or fewer enemy forces in Vietnam because General Westmoreland's command refused a higher count, with Helms yielding to pressure. | Partly supported | Carver's telegram reports that MACV 'has given instruction tantamount to direct order that VC strength total will not exceed 300,000 ceiling' and refused to quantify irregulars, and it frames the issue as one of press criticism and public relations. The words Weiner quotes from Carver ('politically unacceptable total of over 400,000') do not appear in that telegram, so the quotation needs a source Weiner cites in his notes (not checked). The telegram does not show Helms's response. |
| Over the cold war the CIA spied on citizens, tapped telephones, opened first-class mail and conspired to commit murder 'on orders from the White House'. | Partly supported | The Committee found plots against Castro and Lumumba conceived by U.S. officials, and U.S. weapons given to the group that killed Trujillo. It found it 'difficult to be certain at what levels assassination activity was known and authorized' and states it could not make a finding that the plots were authorized by the Presidents. Domestic surveillance is documented: nearly a quarter of a million first-class letters opened 1953-1973, files on about 7,200 Americans under Operation CHAOS (Book II). On the Trujillo case Weiner writes that the killers shot him 'two weeks' after the January 1961 arms decisions and after a Kennedy-approved cable; the Committee dates the cable 29 May 1961 and the killing 30 May 1961, and records that Consul General Dearborn's reading of the cable ('we don't care if the Dominicans assassinate Trujillo') was his own interpretation. The Committee found conflicting evidence on whether the weapons supplied were knowingly given for use in the assassination. Record partly withheld. The records that would show who authorized the assassination plots are partly gone: the Church Committee's Interim Report (S. Rep. 94-465) records that CIA Director Helms ordered the working papers of the CIA's 1967 Inspector General study of the Castro, Trujillo and Diem allegations destroyed, that William Harvey's notes of White House 'urging' were destroyed, that the State Department cabled the Dominican station on 31 May 1961 to destroy all records of contacts with the dissidents, and that the Nixon files were not made available because of a court order. |
| James Schlesinger's May 1973 order produced a CIA compilation of crimes that 'eventually ran to 693 potential violations', including plots against Castro, domestic surveillance, and drug tests on unwitting subjects. | Partly supported | The briefing confirms the 9 May 1973 Schlesinger directive and the contents (assassination plots against Castro, Lumumba and Trujillo, mail opening, wiretaps on columnists, behavior modification experiments on 'unwitting' citizens). It describes the file as a '693-page' compilation per Director Hayden's announcement, and the copy delivered to the Archive as 702 pages, so '693' counts pages in that source, not violations. Record partly withheld. The 1973 compilation is public only in a 2007 version the National Security Archive calls 'substantially-excised' and more heavily excised than the 1977 release, after a FOIA request the Archive says was 15 years old (EBB 222 and EBB 435), and the MKULTRA files themselves were believed destroyed in 1973 on the orders of the retiring Office of Technical Service chief with the authorization of the DCI (Adm. Turner, joint hearing, 3 August 1977, p. 8). |
| In November 1986 CIA officials concealed from Congress that the CIA had flown arms to Iran in November 1985. Deputy Director McMahon had required a backdated presidential finding for that flight. | Supported by the record | Walsh finds 'a concerted effort by CIA officials to withhold information from or lie to Congress about the November 1985 shipment' and that 'George knew that his statement was false' in his 18 November 1986 briefing. He records that CIA general counsel Sporkin 'drafted the retroactive Finding'. Clair George was convicted on 9 December 1992 of two felony counts of false statements and perjury before Congress and pardoned on 24 December 1992. |
| William Casey was 'ultimately responsible' for the arms-for-hostages deal, kept using Ghorbanifar after being told he was a con man, and President Reagan then told the nation on 13 November 1986 that no weapons were traded for hostages, 'lied to protect the covert operations of the CIA'. | Partly supported | Walsh finds Casey 'was at least instrumental in setting up a situation which made the diversion possible' and 'concealed the U.S. role in the 1985 TOW and HAWK shipments', but obtained 'no documentary evidence showing Casey knew about or approved the diversion'. Reagan said on 13 November 1986 'We did not--repeat--did not trade weapons or anything else for hostages' and on 4 March 1987 'My heart and my best intentions still tell me that's true, but the facts and the evidence tell me it is not.' Walsh concluded that Reagan's conduct 'fell well short of criminality which could be successfully prosecuted'. Whether Reagan knowingly lied is not decided in the sections read. Record partly withheld. Walsh's Final Report, ch. 15, records Poindexter's testimony that he destroyed the signed December 1985 Finding on 21 November 1986, records that no copy of the signed Finding was located and that Reagan did not recall signing it, and records that Independent Counsel could not corroborate North's testimony that Casey told him to destroy the ledger of the contra fund; Casey resigned on 29 January 1987 'incapacitated by a brain tumor'. |
| The CIA's October 2002 National Intelligence Estimate asserted that Iraq had continuing weapons-of-mass-destruction programs, and the assertion was wrong. | Supported by the record | The estimate states 'We judge that Iraq has continued its weapons of mass destruction (WMD) programs' and 'Baghdad has chemical and biological weapons'. The Senate Select Committee on Intelligence's first conclusion is that most major key judgments were overstated or not supported by the intelligence, and the Silberman-Robb Commission wrote that the Intelligence Community 'was dead wrong in almost all of its pre-war judgments'. Weiner's quotation of the nuclear sentence stops at 'within several months'; the estimate reads 'within several months to a year'. |
| The CIA's Iraq reporting was 'not a selective use of intelligence' and 'not cherry-picking': Colin Powell's February 2003 UN presentation was the best intelligence the agency had to offer, delivered to a president the agency 'desperately sought' to please by telling him what he wanted to hear. | Partly supported | The Silberman-Robb Commission quotes analysts who 'universally agreed that in no instance did political pressure cause them to skew or alter any of their analytical judgments', and the Senate Committee's Conclusion 83 reports no evidence that officials 'attempted to coerce, influence or pressure analysts'. The Commission does find the daily briefs 'more alarmist and less nuanced' than the estimate and 'seemed to be "selling" intelligence' to the president, which Weiner quotes accurately at. The Phase II report found several public statements by administration officials not supported by the intelligence, which bears on Weiner's description of what officials said but not on the estimate itself. |
| The 6 August 2001 President's Daily Brief, headed 'Bin Ladin Determined to Strike in US', was 'a very weak piece of reporting': 'The freshest intelligence in it dated from 1999. It was a work of history, not a news bulletin.' | Contradicted by the record | The Commission prints the brief. It includes 'FBI information since that time indicates patterns of suspicious activity in this country consistent with preparations for hijackings or other types of attacks, including recent surveillance of federal buildings in New York', about 70 FBI full field investigations, and a May call to the embassy in the UAE reporting Bin Ladin supporters in the U.S. planning attacks with explosives. The Commission also states it was the 36th brief item that year on Bin Ladin or al Qaeda and the first devoted to the possibility of an attack in the United States. The title and date match Weiner's. |
| President Bush's 17 September 2001 directive 'set no limits on what the agency could do' and was the foundation for secret prisons where the CIA used torture; the CIA threw 'hundreds of suspects into secret jails in Afghanistan, Thailand, Poland, and inside' Guantanamo. | Partly supported | Finding #11 states the 17 September 2001 Memorandum of Notification granted authority to 'covertly capture and detain individuals "posing a continuing, serious threat of violence or death to U.S. persons and interests"' and 'made no reference to interrogations or coercive interrogation techniques', which does not match 'no limits'. Finding #15 states the CIA 'detained at least 119 individuals, of whom at least 39 were subjected to the CIA's enhanced interrogation techniques', and that at least 26 of the 119 were wrongfully held. Finding #3 states the interrogations 'were brutal and far worse than the CIA represented'. The book's 'hundreds' of CIA-held suspects exceeds the Committee's count, which covers CIA custody and excludes military detainees at Guantanamo. Record partly withheld. The Senate Committee's 6,700-page study 'remains classified' and only the Findings and Conclusions and Executive Summary were released, with CIA-proposed redactions and 'significant alterations' negotiated with the CIA and White House (S. Rpt. 113-288, Foreword pp. 5-7 and note 6, PDF p. 39); the CIA also destroyed the interrogation videotapes of Abu Zubaydah and al-Nashiri on 9 November 2005 (same report, PDF p. 473). |
What the record shows that it leaves out
- The 5 January 1961 NSC minutes record Eisenhower saying 'a strong central position with respect to intelligence was necessary' and that the function of the CIA was to acquire strategic intelligence. The title scene omits this (FRUS 1961-63 vol. XXV, doc. 80).
- Church Committee Interim Report: no evidence that U.S. officials favored the killing of Diem and none that the United States was involved in Lumumba's killing (S. Rep. 94-465, Summary of Findings). Weiner quotes Johnson's remark that the United States 'assassinated' Diem at shelf copy without these findings.
- Silberman-Robb Commission and Senate Committee Conclusion 83: analysts reported no political pressure on their Iraq judgments. Weiner writes that the agency told the president 'what he wanted to hear' and omits the analysts' testimony.
- 9/11 Commission Report, pp. 261-262: the 6 August brief cited FBI findings consistent with hijacking preparations, about 70 FBI full field investigations and a May report of Bin Ladin supporters planning an explosives attack in the U.S. Weiner's description of the brief omits this content.
- Church Committee Book I: about 50 American journalists or media employees among CIA assets, and more than a dozen U.S. news organizations and publishers that provided cover for agents abroad.
- CIA analytic warnings that Dujmovic cites from declassified records: 1 September and 30 September 1950 reports to Truman that Chinese intervention in Korea was possible, and the 12 October 1950 report that placed full-scale intervention as 'a continuing possibility'; the 1967 Arab-Israeli war assessment (Robarge, Studies in Intelligence 49 no. 1, 2005). Weiner's treatment of these cases is described by Dujmovic as one-sided.
- Records published after the book's 2007 date that bear on its subjects: the Senate Select Committee on Intelligence study of 2014 (119 CIA detainees, 39 subjected to enhanced techniques, 26 wrongfully held; finding that much information the CIA gave the media, Congress, the Justice Department and the White House about the program was inaccurate), and the CIA's 2013 release acknowledging its role in the 1953 Iran coup under the TPAJAX name.
Documented efforts and withheld records
Each item is a documented act, with its source: an act against the work, its author or its sources, or the withholding of records on its subject. The list states what was done and does not state why.
- 1967 (found by the Committee in 1975-76). CIA Director Helms ordered the working papers of the CIA Inspector General's 1967 study of the Castro, Trujillo and Diem assassination allegations destroyed when the study was complete (memorandum for the record, 23 May 1967, as cited by the Church Committee). Church Committee, Alleged Assassination Plots Involving Foreign Leaders, S. Rep. 94-465, introduction (PDF p. 15)
- 31 May 1961 (cable); Harvey testimony 25 June 1975. The State Department cabled the CIA station in the Dominican Republic to destroy all records concerning contacts with dissidents, except the contingency plans and the 29 May 1961 cable. The Committee also records that Harvey's notes of White House urgings to develop an assassination capability had been destroyed. Church Committee, S. Rep. 94-465, Trujillo and Executive Action sections (PDF pp. 225 and following)
- Report of 26 April 1976. The Church Committee reports that the CIA denied it access to the names of its media agents and of the organizations involved, to records on media relationships before 1967 (other than screened files), and to certain titles and authors of propaganda books published after 1967; the Committee saw summaries and about 20 files screened by the Agency with names removed by agreement. It states it has 'far from the full picture'. Church Committee Book I (S. Rep. 94-755), Covert Relationships with the United States Media (PDF pp. 184 and 198) and Introduction (PDF p. 16)
- Through February 1976. CIA maintained covert relationships with about 50 American journalists or employees of U.S. media organizations until it announced a new policy in February 1976 (documented by the Church Committee from summaries the CIA supplied). Church Committee Book I, Covert Relationships with the United States Media, p. 192
- 1973 (destruction); 3 August 1977 (testimony). The retiring head of the Office of Technical Service, with the authorization of DCI Helms, destroyed the MKULTRA behavioral-modification files in 1973; Adm. Turner told the Senate it had been believed all such files were destroyed until a set of financial records was found in 1977. Joint hearing, Project MKULTRA, statement of Adm. Stansfield Turner, p. 8 (exhibit Exhibit 226)
- 21-26 June 2007. The CIA released the 'Family Jewels' compilation in 2007 in a version the National Security Archive describes as substantially excised, in response to a National Security Archive FOIA request the Archive says was 15 years old. National Security Archive, EBB 222 (redaction and page counts) and EBB 435 essay (age of the request)
- 1999 lawsuit; April 2000 leak; August 2013 release. In response to a 1999 National Security Archive FOIA lawsuit, the CIA cited foreign government information and foreign relations and released all but a single sentence of Donald Wilber's 200-page internal history of the 1953 Iran coup. The New York Times obtained the history by a leak and posted it in April 2000. The CIA released a version in 2013 that states the agency planned and helped implement the coup. The Archive's 2013 essay also reports the State Department's Iran retrospective volume of Foreign Relations of the United States was still unpublished and describes CIA and British objections. National Security Archive, EBB 435, 'CIA Confirms Role in 1953 Iran Coup' and the FRUS essay on the page
- 1979 lawsuit; 23 May 1997 release. The CIA answered FOIA requests on its 1954 Guatemala operation with 'neither confirm nor deny'. A 1979 FOIA lawsuit led to the assassination-planning records being gathered and withheld on national security grounds. On 23 May 1997 the CIA released about 1,400 pages of an estimated 100,000 or more, and removed the 58 names on its list of targets and other names. National Security Archive, EBB 4, 'CIA and Guatemala Assassination Proposals' and release notes
- April 2004 request; 31 October 2005 report; 1 December 2005 release. National Security Agency officials tried to block release of the Hanyok Tonkin study. The National Security Archive says release came on 1 December 2005 largely through FOIA requester Matthew Aid, who requested the study in April 2004 and told the New York Times of the attempt; the Times reported on 31 October 2005 that NSA had kept the study secret since 2001. National Security Archive EBB 132 release note; Scott Shane, New York Times (via History News Network)
- 21 November 1986; October-November 1986. Poindexter destroyed the signed December 1985 Iran arms Finding on 21 November 1986, and North testified he destroyed the ledger of the contra fund in October or November 1986 at Casey's direction (Independent Counsel could not corroborate the Casey instruction). Walsh, Final Report of the Independent Counsel, ch. 15 (Casey)
- 9 November 2005; disclosed December 2007; Durham closed 8 November 2010. The CIA destroyed 92 videotapes of the interrogations of Abu Zubaydah and al-Nashiri on 9 November 2005, the day after the Senate rejected an amendment to create a detainee commission. The destruction became public in December 2007. Special prosecutor John Durham closed his investigation on 8 November 2010 without recommending charges. Senate Select Committee on Intelligence study, S. Rpt. 113-288 (destruction and its role in starting the study); the 92-tape count, Durham outcome and press references from Wikipedia, 'CIA interrogation tapes destruction' (secondary)
- 1 August 2014; released December 2014. The CIA returned the Senate study's Executive Summary on 1 August 2014 with proposed redactions. The Committee states significant alterations were made to reach a releasable version: dates replaced by general time frames, true names replaced by pseudonyms, then all pseudonyms and some job titles redacted by the executive branch. The full 6,700-page study remains classified. S. Rpt. 113-288, Foreword by Chairman Feinstein and note 6 (PDF pp. 5-7, 39)
- January 2014 searches; summary released 31 July 2014. CIA personnel searched the Senate Intelligence Committee's side of a shared classified network and Senate staff email. The CIA Inspector General's summary found five CIA employees (two lawyers and three computer specialists) accessed the Committee's computers, that a CIA security office reviewed staff email after Director Brennan ordered a halt, that three showed 'a lack of candor', and that the CIA's criminal referral of Senate staff to the Justice Department rested on inaccurate information. Brennan told reporters in March 2014 'We wouldn't do that' and apologized in July 2014. The Justice Department declined charges. Associated Press via Christian Science Monitor, 1 August 2014; The Intercept, 5 January 2015 (Accountability Board chaired by Evan Bayh, reported by the New York Times as not recommending discipline)
- 6 August 2007 (per Wikipedia). The CIA issued a press release when the book appeared saying Weiner had the sources to write a balanced history but his bias 'overwhelms his scholarship'. CIA press release, as reported in Wikipedia, 'Legacy of Ashes'
- September 2007 (vol. 51 no. 3). CIA historian Nicholas Dujmovic reviewed the book in the CIA's journal Studies in Intelligence, calling it 'a 600-page op-ed piece masquerading as serious history' and factually unreliable, and the title scene invented. Studies in Intelligence, Center for the Study of Intelligence, CIA
Published responses
- Nicholas Dujmovic, CIA historian, Studies in Intelligence 51 no. 3 (2007): Calls the book 'a 600-page op-ed piece masquerading as serious history' and says it is 'factually unreliable'. Says the title scene is 'an invented dialog' that 'never happened': Dulles spoke at the 12 January 1961 NSC meeting against a Defense proposal to separate the director from CIA management, and Eisenhower's 'legacy of ashes' remark was made a week earlier on 5 January about military intelligence; the words 'preposterous' and 'madness' are not in the record. Lists factual errors: the 1949 CIA Act did not grant covert action authority; the CIA never used military ranks ('CIA colonel', p. 88); Hayden was not the first active-duty officer to lead the agency since the early 1950s (Turner, 1977); John McCone was never a deputy secretary of defense; the Afghan mujahedin the CIA supported are distinct from the Arabs it did not support; the Downey and Fecteau flight of 1952 was a document pickup, not a rescue. Says Weiner dated the 500-ICBM projection to 1960 when it was made in 1957, and that a 1960 NIE estimated about 10 operational Soviet ICBMs. Says Weiner omitted that the 'sixteen words' on Iraqi uranium were attributed to British intelligence and that the CIA had removed the assertion from earlier speeches. Says Weiner quoted Abbot Smith on Soviet analysis without the next paragraph of the same document, and misread Michael Haas on Korea. Credits Weiner with accurately chronicling the early OSO and OPC disorder and says his 1990s chapters 'seem accurate and useful'. Predicts a historian will replace the book.
Where it connects
- Covert Action and Accountability: Iran 1953, Guatemala, the Church Committee findings and Iran-Contra are the records the chapter reads, and the destroyed files, excised releases and FOIA suits above are the chapter's record of what was hidden.
- What Was Said in Private: Tonkin, the 1967 enemy-strength dispute, the 6 August 2001 brief and the 2002 Iraq estimate set private findings beside public statements; the Tonkin study release forced by a FOIA request belongs here.
- Managing the Narrative: The Church Committee's count of about 50 CIA relationships with U.S. journalists, the CIA's screening of what the Committee saw, and the CIA's own response to the book bear on how an agency shapes what the public reads about it.