Frederick J. Graboske, Nixon Tapes Archivist, explains why he concluded Stanley Kutler's alterations were deliberate
Preparing transcripts of the Nixon tapes is a difficult task. My staff at the National Archives and I did a number of them. With experience we got our transcription time down from 400 hours per hour of conversation to 100 hours. The results were good, but not perfect. In addition to the relatively poor quality of most of the recordings, we were faced with rendering “natural conversation,” with its repetitions, stuttering, and slurred words. The key phrase in one transcript was “ . . . so Harlow didn’st shustem/shust ’sim (rendered phonetically).” Everyone who listened heard that meaningless word/phrase. The participants had a shared knowledge of the subject, and no one questioned the speaker about what he meant.
On another occasion we had prepared a transcript for Judge Gesell to review in camera. I took the transcript and a copy of the tape to his chambers to play for the Government lawyers and for Nixon’ss counsel, R. Stan Mortenson. At one point Stan interrupted the proceeding and said he had an alternate rendering of one of the sentences: same cadence, different words, different meaning. The Government’ss lawyers agreed, only to have Stan say that this incident proved the suggestibility to the written word of people listening to the tapes. In fact, he said, that was not at all what he had heard and he presented a third version: same cadence, different words, different meaning. It was then that I realized that transcriptions of the Nixon tapes never would be wholly acceptable to men of good willsomeone always would hear some word or phrase differently. These criticisms would cast a pall over the entire transcription and, therefore, I decided against producing a Governmentsponsored set of transcripts.
With this background I approach with sympathy anyone attempting to produce generally acceptable transcripts of the Nixon tapes, such as Dr. Kutler. In his editorial note he stated that he removed extraneous wordsthe repetitions, stutters, etc. He notes also that some may disagree with particular renderings of words. I have no problem with his work on any of these issues. My fundamental disagreement lies with the conflation of portions of two transcripts from two different tapes (Oval Office and WH Telephone), recorded hours apart.
Preparation of a transcript is done by working from the tape logs — the topic outlines of every conversation prepared by the Archives staff. The logs note the conversation number (the tape number plus the position of the conversation on the tape, such as 881-03), the participants, and the start and stop times of the conversation. As in all historical research, context is important; the times listed on the logs provide the sequence for the conversations. Once a conversation, or a portion thereof, has been identified for transcription, someone listens to the tape and produces a rough transcript. Sheer mental fatigue precludes effective transcription more than about 4 or 5 hours per day, so others review the transcript until there is general agreement. The typed document would retain its tape identification information and its pages numbered in sequence. Each transcript is integral.
To conflate 2 transcripts would require literal or electronic cutting and pasting. This is a deliberate act. Of course, one could imagine a scenario in which the physical pages of more than one transcript (say, those for March 16) were scattered on a desk and accidentally merged, despite the conversation identifiers and page numbering. However, I assume that the court reporters who prepared the transcripts provided them in both electronic and physical format, so that a diligent author could check his work with the physical transcripts against the electronic form. This “accident” scenario implies a level of sloppiness on the part of the researcher/author that casts a pall not only over the publication in question but over the entire corpus of his work. I choose not to believe that of Dr. Kutler. He states in his forward that he is “ . . . aware of my responsibility for accuracy” and that “ . . . there is no distortion of the thrust or intent of the passages. “ The conflation of the two transcripts demonstrates that he failed in that responsibility.
By choosing to publish only portions of the Nixon/Dean conversations Dr. Kutler asks us to trust his historical judgment on relative importance. Obviously, the physical constraints of publishing a book of standard length preclude publishing the entire corpus of transcripts, although a CD-ROM could have been includedas Bob Haldeman did with his diaries. The danger here is one of lack of context. To understand Nixon and his actions with regard to Watergate, these conversations should be seen in the context of other conversations he had on these topics, including those which Dr. Kutler chose not to include, such as March 13. A portion selected for publication is torn from the context of the rest of the conversation. This is a problem inherent with such a “highlights of Watergate“ book: the selections could be seen as agenda-driven.
The Watergate tapes are available on-line at nixontapes.org. Researchers should use this site as their primary source rather than relying on the flawed Kutler book.
HNN Hot Topics: The Watergate Transcript Controversy
comments powered by Disqus
Maarja Krusten - 3/7/2009
My old boss, Fred (hi, Fred!) writes, "At 100 hours of staff time per hour of conversation (this is the actual rate at which we transcribed), it would have taken 400,000 staff hours to produce transcripts. That's 200 staff years. My staff varied in size between 6 and 8 people. 8 people would have taken 25 years to transcribe the tapes."
It is true that over time, we reduced the transcription ratio from about 300 staff hours per hour of tape to 100 hours per hour of tape. However, at the time the National Archives first considered how to handle the descriptive phase of processing the Nixon tapes (around 1978 or 1979), we had not yet reduced the time down to 100 hours. Also, I don't believe we had foot pedal controls for our analog equipment when we first did transcription. (Quite comically, I occasionally find my right hand reflexively reaching out for a nonexistent Tandberg open reel recorder's gear shift sometimes when I listen to web posted audio clips for the Nixon tapes and want to listen to a portion again, LOL. No, Maarja, ya gotta click the mouse. There sure are some very well worn grooves in my brain, I guess.)
When we did our first transcripts for court cases, a secretary typed them on a typewriter. NARA had not yet acquired Datapoint word processors, which enabled us to reduce typing time for tape survey logs and the few transcripts we did during the 1980s. Computers came later. I was the tapes archivist who tested the first IBM-type PC at the Nixon Project at the end of 1989, shortly before I left NARA's employ in January 1990.
I do hope that you wandered back to read our comments, Dr. Ritchie. I'd love to hear from you here on HNN, given your expertise with oral history and so forth. While I don't know you personally, I do know and respect greatly your colleague, Dick Baker. Please pass along my regrds to Dick!
Frederick Graboske - 3/5/2009
I must beg to disagree with Dr. Ritchie on his assertion that NARA sidestepped its responsibility to make transcripts of the Nixon tapes. As we discovered when preparing transcripts under court order, key words and phrases could be heard in different ways by listeners. Thus, any transcripts that we prepared would be subject to challenge; they could not be considered authoritative. Further, transcripts are derivative of the primary source materials: the actual recorded conversations. These are what are called "natural conversation", as opposed to the structured conversation of a speech or an oral history interview. Tone and inflection are more important in natural conversation than in other types. Transcription cannot render these.
There also were 2 very practical reason for not producing transcripts: the level of effort required and possible objections by Nixon. At 100 hours of staff time per hour of conversation (this is the actual rate at which we transcribed), it would have taken 400,000 staff hours to produce transcripts. That's 200 staff years. My staff varied in size between 6 and 8 people. 8 people would have taken 25 years to transcribe the tapes. Of course, it's been 31 years since we started, but our plan called for first creating the topic outlines, "tape logs", for the conversations, done by hand (these were the days before computers), so that we could gain intellectual control over the tape contents and respond to legal requests. Given the problems of accurate transcription of the Nixon tapes, we thought it quite likely that Nixon would challenge the release of many, if not most, transcripts. We completed the tape logs and the archival review of the tapes in 1988. Extraneous factors delayed, and continue to delay, the full release of he tapes.
Given the real (accuracy, length of time required) and anticipated (Nixon objections)problems with transcription, we opted to let the tapes speak for themselves. We did this in full understanding that listening to the tapes can be difficult and tedious. We did this in full understanding that NARA had reported to Congress that it could have transcripts ready 3 years after assuming custody and control of the tapes. None of us who worked on the Nixon materials had any part in preparing this report.
The level of detail of the tape logs--25,000 pages--was intended to enable researchers to focus on specific portions of conversations, thereby ameliorating somewhat the problems inherent in listening to the tapes.
With regard to my critique of Dr. Kutler's work, I did not criticize his renderings of the portions of conversations he selected for transcription. I did not criticize those selections. I did take issue with his interpolation of a portion of one conversation into another. This act either was deliberate or an example of very slipshod work. Either way, it casts doubt on the general accuracy of these transcripts and reinforces my decision 30 years ago that the tapes themselves should be consulted as the primary source materials.
Maarja Krusten - 3/5/2009
Oops, NHF should be NHJ. The judge hearing the case in District Court ended up by Norma Holloway Johnson. Initially the case was assigned to Gerhard Gesell, later to another judge (Thomas Penfield Jackson) then finally to Norma Holloway Jackson. As a result, deposition transcripts from 1992 variously carry the initials GAG, TPJ, and HHJ in the case number.
Maarja Krusten - 3/5/2009
Here is some additional information on NARA’s corporate decision not to produce “official” transcripts of the Nixon tapes except when required to do so under subpoena for litigation.
Dr. Kutler filed a lawsuit against the National Archives in March 1992 in an effort to gain access to the Nixon tapes (Kutler v. Wilson, Civ. A. 92-0062-NHF). A number of officials testified under oath in the case, including Assistant Archivist for Presidential John T. Fawcett, former NARA Nixon Project Supervisory Archivist Frederick J. Graboske, and others. I was among those called to testify. Some of the witnesses were asked about the nature of the archival processing that had taken place with the Nixon tapes since NARA took physical custody of them in August 1977. In his 1992 testimony, the most senior official among the witnesses, Mr. Fawcett, explained that NARA had decided early on in its work with the Nixon tapes not to do transcription. Testimony in the Kutler case suggested that such would have been deemed to be against archival principles.
I worked at NARA at the time that decision was made and can explain the reasoning. We at NARA believed it was the audio recording that was the record. Mr. Graboske illustrates well in his article why reasonable people may disagree as to what is being said on the tapes. The 1988 article I cited above provides the perspective of his supervisor, the director of the Nixon Project, as to why NARA chose not to transcribe the tapes. Clearly, NARA management had concerns (1) about the time it would take to transcribe all the tapes (2) and the feasibility of a federal agency producing official transcripts for which it could offer no guarantee of 100% accuracy. When we started working with them, we estimated there were 4,000 hours of taped conversations. (We later revised that to 3,700.) Multiply 4,000 hours by 300 hours per hour of conversation and it is clear that we could have spent decades transcribing the tapes, while researchers had to wait for access to them. Keep in mind that we worked under complicated regulations that took into account the needs of multiple stakeholders and that our work proceeded in carefully planned phases.
The March 1975 Report to Congress on Title I of the Presidential Recordings and Materials Preservation Act provides an initial glimpse at our plans, as do hearings held then. The Kutler litigation also covered what these phases entailed. I’m happy to summarize them for you and explain how a Reagan-era Reduction in Force (RIF) affected the plans set out in the 1975 GSA Report to Congress.
Here’s how the archival processing played out. Phase I (Preservation) involved rewinding and duplicating the original tape reels, a process that began in March 1978 and concluded in August 1978. Phase II (Description) involved identifying what was on the then estimated 4,000 hours of tapes. Keep in mind, we largely did not know what was on the tapes when we moved them out of the White House in August 1977. The Special Prosecutor only examined a small portion of them during Watergate investigations. NARA Phase II involved preparation of subject logs, a process which ran from 1978 through 1981 (with some revision taking place in 1986). Responding to subpoenas, we also produced transcripts of a few excerpts of tapes around 1978-1979.
As it happens, I helped develop the tape subject log format we decided to use to describe the tapes. Along with my superiors, Mr. Graboske as well as the director of the Nixon Project, I take responsibility for our corporate decision to log rather than to transcribe the tapes. Rather than making researchers wait for decades while we tried to produce transcripts with, at best, "99% accuracy," we decided to create a descriptive outline of the topics covered in the conversations. We called this the tape survey log. Keep in mind, our goal was to start releasing information about the Nixon presidency as soon as possible. We also were trying to be responsive to a law that required us to release "the full truth" about Watergate at "the earliest reasonable date."
You can find examples of our logs at
As Phase II wound down, we segued into Phase III (Review). This lasted until August 1987. It involved listening again to all 3,700 hours of the tapes to determine what could be released and what required restriction.
All of this work took place with far fewer staff resources than anticipated in the 1975 Report to Congress on Title I. Where the Report projected a staff for the National Archives’ Nixon Presidential Materials Project of some 100 employees, the highest number we ever had on board was about 40. In 1981-1982, we were hit hard by the Reagan RIF. As you know, RIFs then largely were run on a “last hired, first fired” principle. As a new Project within the National Archives which just had staffed up, we ended up losing 50% of our staff during the RIF. Budget constraints continued to affect NARA and we at the Nixon Project never fully recovered from the Reagan RIF. For much of the 1980s, work with the tapes was carried out by about 6-8 tape reviewers, including the supervisor and I (senior tapes archivist). As a functional unit, NARA’s Nixon Presidential Project never came close to having the 100 employees envisioned in the 1975 Report to Congress on Title I.
In addition to supporting what Mr. Graboske noted, I hope this helps clarify that the decision not to transcribe the tapes was a corporate decision on the part of the National Archives. As public servants, we had to balance competing obligations and did the best we could under the circumstances.
Posted from home, natch
Maarja Krusten - 3/4/2009
I once worked at NARA with and for Mr. Graboske, a former federal official whom I respect greatly. You may not be familiar with the record regarding NARA and our decision not to transcribe the Nixon tapes. Please see
which provides some needed background information on this.
The issues here are complicated and need to be approaced with care.
Former National Archives' Nixon tapes archivist
Posted by Smartphone on personal time (lunch break)
Donald A. Ritchie - 3/4/2009
It is hardly edifying for Frederick Graboske to offer an excuse for why the National Archives sidestepped its responsibility to transcribe the Nixon tapes and then to question the motives of a historian who used his own resources to do what the Archives should have done. The French have a good expression for this: qui s'excuse accuse.
- Did a historian who said he’s a victim of McCarthyism get the story wrong?
- Stephanie Coontz’s work on the history of marriage cited by the Supreme Court.
- NYT History Book Reviews: Who Got Noticed this Week?
- David Hackett Fischer wins $100,000 prize for lifetime achievement in military writing