Errors Still Afflict the Transcripts of the Kennedy Presidential Recordings

tags: JFK, presidential libraries, Kennedys, transcripts

Mr. Stern, historian at the Kennedy Library from 1977 to 1999, is the author of Averting 'The Final Failure': John F. Kennedy and the Secret Cuban Missile Crisis Meetings , recently published in the Stanford University Press Nuclear Age Series.

Everything has a history including the writing of history itself. It has now been four years since I first discovered important transcription errors in Ernest May and Philip Zelikow, eds., The Kennedy Tapes: Inside the White House During the Cuban Missile Crisis (Harvard University Press, 1997). That discovery was entirely serendipitous. I had just begun relistening to the tapes as I began working on my own narrative account of the ExComm meetings and had already decided to use JFK's October 18 th description of nuclear war as"the final failure" in the title of my book. However, when I looked through the May-Zelikow transcript, I instead found the words"the prime failure." Naturally enough, I assumed the editors were right because they had better technology. However, when I listened again, I found that JFK had unquestionably said"the final failure." As a result, I listened to the entire tape and discovered so many errors that I finally decided to write an article ("What JFK Really Said," Atlantic Monthly , May 2000).

The editors responded by claiming to be"bemused" and"wryly gratified" that I had found only seventeen transcription errors in more than 200,000 words of transcripts and also concluded that"none of these amendments is very important" ( Atlantic Monthly , August 2000). In response, I published a second article identifying many more errors ("Source Material : The 1997 Published Transcripts of the JFK Cuban Missile Crisis Tapes: Too Good to Be True?" Presidential Studies Quarterly , September 2000). The editors later acknowledged that their use of court reporters to prepare the initial transcripts had been a mistake and that some of their sound enhancement technology had been disappointing (Presidential Studies Quarterly , December 2000).

There the matter rested until the publication of Philip Zelikow, Ernest May, and Timothy Naftali, eds., The Presidential Recordings: John F. Kennedy: Volumes 1-3, The Great Crises ( Miller Center-W.W. Norton, 2001). My review of these revised and improved transcripts (Reviews in American History , December 2002) concluded that the new missile crisis transcripts (about 35 percent of the three volumes) were substantially more accurate than the 1997 version, but still contained far too many significant errors.

In February 2003, Philip Zelikow and I both spoke at a conference on presidential tapes at the John F. Kennedy Library. In his prepared remarks, and later in the question and answer segment, Professor Zelikow publicly thanked me for my contributions to this complex task and invited further scholarly criticism. He also announced that the Miller Center had, that very morning, activated a new website, whitehousetapes.org, to allow scholars to make corrections to the published transcripts and to download corrected transcripts from a"multi-media errata sheet." Professor Naftali later explained that this website would include"full attribution" to scholars making these corrections.

Five months after the conference, my book, Averting"The Final Failure," John F. Kennedy and the Secret Cuban Missile Crisis Meetings, the first narrative account of the ExComm meetings, was published by Stanford University Press. The book includes an Appendix citing some 100 key errors in the new Miller Center-Norton missile crisis transcripts (divided into three categories: remarks transcribed as"[unclear ]" which are, in fact, discernable; words, phrases, sentences or speakers missing altogether from the transcripts; mistranscriptions which alter the substance or interpretation of the historical record). Many additional differences appear in the body of the book but are not highlighted separately in the Appendix.

Readers should also understand that presidential tapes scholars may even disagree on the identification of voices and all of us can and should learn from the following examples. When Professors May and Zelikow and I talked briefly at a 1997 Kennedy Library conference, we disagreed on several important identifications. I argued that the member of the Joint Chiefs who complained on October 19 that the military was being"screwed" by JFK's blockade decision was General Curtis LeMay. They insisted it was General David Shoup. I went back and listened again and concluded that they were indeed right: it is definitely Shoup. However, we also disagreed on identifying one of the JCS officers who briefed JFK on October 29. They suspected that it was Admiral Robert Dennison, commander of the Atlantic Fleet, and seemed skeptical of my claim that it was Admiral George Anderson, Chief of Naval Operations. I had never heard Dennison's voice but contended that Anderson had also spoken at the October 19 meeting and the voices at both meetings were the same. We agreed to disagree.

However, when I was writing my book in 2001, I discovered that the 1997 Harvard Press editors had entirely missed a revealing exchange in their transcript of the October 29 briefing. (The Miller Center-Norton edition only goes up to October 28.) The missing exchange is as follows: JFK:"I see you're on the cover of Time, admiral.""Sir," the admiral replied,"I haven't read the article yet.""I'm sure they'll be kinder to you," Kennedy observed,"than they are yet to me." I checked the Time Magazine website and confirmed that the cover subject on October 28, 1962 was indeed Admiral George Anderson. That kind of definitive identification is certainly the exception rather than the rule in this kind of research.

In any case, a year has now passed since the Kennedy Library tapes conference. The Miller Center website does cite, under"reviews" and"research," my three articles on the published transcripts, as well as my book. Nonetheless, it seems reasonable and fair to ask when the promised"multi-media errata sheet" will appear on this new website. It is entirely possible that Miller Center scholars disagree with some of my transcriptions, and in some cases, they may of course be right. But there is no doubt in my mind that the great majority of my corrections will stand up to scrutiny, as suggested by Professor Zelikow's gracious decision, in his words,"to pay tribute to the work of people like Sheldon Stern," at the 2003 JFK Library conference. As of today, however, the online transcripts are simply pdf versions of the 2001 Miller Center-Norton volumes: no corrections have been made. Historians also deserve some explanation for why the Miller Center's multiple listener-back up system ("every transcript has benefited from at least four listeners," using"the team method,""a special kind of 'peer review'" and"the best technology that the project can afford") has still resulted in so many errors.

Let me cite just one out of hundreds of examples to make clear to readers, especially historians, that these changes can be essential to our understanding of the missile crisis itself. At the Saturday evening meeting on October 27, when the situation looked extremely grim and no one imagined that Khrushchev would pull back from the brink the following morning, Secretary of Defense Robert McNamara urged the president to turn up the heat on the Soviets. McNamara, in both the 1997 Harvard Press and the 2001 Miller Center-Norton transcripts, says,"I would say only that we ought to keep some kind of pressure on tonight and tomorrow night that indicates we're firm. Now [added in the 2001 transcript] if we call off these air strikes tonight, I think that settles that." (italics added; 1997, p. 612; 2001, vol. 3, p. 492). In fact, these words make no sense historically or logically: first, there were no air strikes scheduled for that Saturday night; second, how would calling off air strikes increase the pressure and demonstrate firmness? McNamara was instead recommending the call up of 24 air reserve squadrons. He actually said:"I would say only that we ought to keep some kind of pressure on tonight and tomorrow night that indicates we're firm. Now if we call up these air squadrons tonight, I think that settles that" (italics added). It is hard to imagine that any historian would contend that this distinction is not important.

The incorporation of these kinds of changes on whitehousetapes.org would demonstrate that the Miller Center, as Professor Zelikow publicly stated, is committed to making corrected transcripts available for downloading and ready to welcome, with full attribution, the contributions of the wider community of presidential tapes scholars.


In advance of the publication of this article, HNN forwarded a copy to Philip Zelikow for a response. We received the following message.

We do indeed plan to compile and post the multimedia 'errata' sheet, as promised. Transcription is a challenging process. I think those who have studied the tapes and the Miller Center volumes, noticing our scene-setters and annotations, can judge the extent to which we have advanced the understanding of these historically vital records.

The completion of the 'errata' has been delayed in part by my own preoccupation with completing the work of the 9/11 Commission, which I direct. Once that task is completed, later this year, my colleagues and I look forward to doing this work to make this scholarship a living project, benefiting from collegial suggestions for improvement.

comments powered by Disqus

More Comments:

Robert Love - 3/18/2004

A minor contribution. Many years ago, when only a few of the tapes had been released, someone - perhaps a graduate student - working at the National Security Archive in Washington apparently learned of my interest in the U. S. Navy's role in the crisis and provided me with a copy of the tapes for one afternoon's meetings - or some portion thereof. I cannot recall the exact date. One of my Honor's students took on the project of transcribing and annotating the transcription as a semester project, but he was soon frustrated by (a) the poor sound quality and (b) his inability to identify the speakers. Being a "prior enlisted" submariner, he turned to a buddy who then worked at a sound lab and who got permission from his CO to "enhance" the sound on the tape with equipment perhaps still unavailable on the commercial market. Having interviewed many of the flag and general officers involved, I could identify some of the speakers, but scarcely all. Nonetheless, the enhanced tape contained so many inaudible sounds that the resulting transciption seemed to me to be of limited value except as a learning tool for my student, who now fully understood the frustrations of historical research. My student turned both his transcription and the tapes back to the fellow at the Archive office and we never heard from them again. Given this history I am naturally unsurprised at the discrepancies noted by Mr. Stern and others.

Sheldon M. Stern - 3/17/2004

I agree completely with Mr. Salmanson. One cannot assume that a trained historian can do this work--some can and some can't. Several top historians have told me that when they listened to the JFK tapes they could not make out anything. It's much like skill at foreign languages--some have it and some don't and it is not determined by intelligence or education. In any case, the "air strikes" vs "air squadrons" example should have been flagged simply because it makes no sense in the context of the missile crisis on that day.

Gabrielle Daniels - 3/17/2004

The Miller Center has a large staff working on its transcripts and although we all praise the 9-11 Commission's work, to use the commission as an excuse not to make changes to acknowledged errors is just wrong. Delegate!

David Lion Salmanson - 3/17/2004

Except for interviews which I conducted, recorded, and transcirbed myself, I've noticed a definte problem with transcriptions. First, if they are not done immediately after the fact, valuable data is corrupted by faulty memory, tape deterioration, confusing with other interviews etc. Second, when somebody else transcribes an interview that they did not conduct, the transcript is littered with errors. I've seen transcripts that had a dozen errors on one page of typed text that had a lot of white space. Third, transcription is a talent (and one I am not particularly good at it if the interivew involves more than one person and myself). Just as some people are good at listening to a record and "hearing" production, particular instruments etc. when other people only "hear" songs in their totality, transcribers not only need to be very familiar with the subject matter at hand (so as to avoid errors like the air strikes one cited above) but be able to "hear" the tapes: voice recognition, picking out simultaneous speech, filtering out backround noise and dozens of other variables. Generally, we do not have a system for evaluating transcribers. On big studies, the work tends to be contracted out or assigned to students. And quite frankly, you get what you pay for.

Sheldon M. Stern - 3/15/2004

I fully understand that Professor Zelikow is preoccupied with the work of the 9/11 Commission. Perhaps, in the interim, another Miller Center scholar could begin updating the corrected transcripts. In any case, we all look forward to the appearance of the 'errata' sheet and the fulfillment of our common goal of making these transcripts "a living project, benefiting from collegial suggestions for improvement."