FRUS? Pitfalls with the Nixon Tapes
and How to Avoid Them
views presented here do not necessarily reflect those of the United States
Government or the U.S. Department of State. Richard
A. Moss served as a graduate student
intern in the Office of the Historian at the Department of State from 2001-2006
and returned as a contract historian during the summer of 2007 to overhaul the
quality control system for the Nixon tapes transcripts in the Foreign
Relations of the United States series.
This included the recently-released American Republics volume.
NEWS REPORTS: STATE DEPT OFFERS
NEW CAVEAT ON NIXON TAPES
July 20, 2009, Secrecy
that the Office
of the Historian at the U.S. State Department
had quietly revised the preface
to latest volume of the Foreign Relations
of the United States to be published, an electronic-only volume on “American
The new preface states that tape transcripts are merely “interpretations” of
the tapes themselves, which are the original “documents.” Steven Aftergood,
the author of the story and head of the Federation of American Scientists
Project on Secrecy, also detailed a number of embarrassing
in a volume on the Soviet
published in December 2006.
revised disclaimer was long overdue. In the late 1970s, the National Archives
and Records Administration (NARA) decided that the recordings were the record
and that its transcripts prepared for a number of court cases should not be
regarded as 100% accurate. The State Department could have learned from NARA’s
example. Since day one, nixontapes.org
has urged readers to trust their own ears and to listen to the audio themselves.
That is the point of the website: to make the audio easily accessible and to
thereby empower researchers.
transcription can be prone to error, even under the best of circumstances.
Unfortunately, the problems highlighted by Mr. Aftergood have less to do with
differences of “interpretation” and more to with the work environment
described in detail by a recent
report by the Office of the Inspector General (OIG).
The OIG report scarcely mentioned the tapes project—something of an oddity
inasmuch as nearly every volume that covers the Nixon administration includes
transcripts. Furthermore, disagreements over the tapes were one of the
underlying sources of tension in the office.
simple fact is that quiet, internal warnings were ignored, with the result that
the entire quality control process had broken down by 2007, thereby jeopardizing
the reliability of the Foreign Relations series.
Failures in the quality control process also enabled the publication of a number
of flawed transcripts in the official documentary record of U.S. foreign policy.
article documents in greater detail the problems exposed by Mr. Aftergood and
includes digital audio of complete conversations so that readers can decide for
themselves. More importantly, we explore some of the techniques and resources
currently available to improve the quality of transcripts in the future
As the official documentary record of U.S. foreign policy since 1861, the Foreign
Relations should be the gold standard. The transcripts are the
product of more than just individual historians working in isolation. They are,
rather, the result of a team of scholars working together (ideally) to identify,
transcribe, and review pertinent material—at the cost of hundreds of thousands
of taxpayer dollars invested in human resources, hardware, software, etc.
Office of the Historian has suffered from a number of problems in recent years,
as documented in the OIG’s recent report. Following the report, the Department
of State acted swiftly at the highest levels to develop some corrective
measures. Some of the recommendations have already been implemented.
Unfortunately, other recommendations are still in process, while others remain
in limbo. The fact is that the relatively short-term problems described by the
OIG will play out on the pages of Foreign Relations for years to come.
are also problems beyond those documented in the OIG report. For example, in his
capacity as an Acting Division Chief, the editor responsible for the mistakes
detailed by Mr. Aftergood is apparently overseeing the Nixon Tapes project as of
June 12, 2009. Since the new disclaimer notes that, “the editor has made every
effort to verify the accuracy of the transcripts,” it is clear where the fault
lies. We can only hope that poor judgment and lack of attention to detail
demonstrated by this compiler does not resurface in future FRUS
2004 and the summer of 2007, quality control standards were inconsistently
applied to Nixon tape transcripts. While some published transcripts are
relatively sound, others have embarrassing errors that could have been easily
avoided with proper review. After all, when we think of the North Vietnamese
peace negotiators Xuan Thuy and Le Duc Tho, we usually think of them as being
tenacious, rather than cuddling,
While the new disclaimer is a belated step
in the right direction, several troubling problems persist. The new preface
states, “Through the use of digital audio and other advances in technology,
the Office of the Historian has been able to enhance the tape recordings and
over time produce more accurate transcripts.” This statement is misleading.
reference to “the use of technology…to enhance” the audio is actually a
reference to noise-reduction software. The
disclaimer implies the Nixon Tapes Group at the Office of the Historian made
widespread use of such noise reduction software, which was simply not the case.
Anybody who has ever used noise-reduction software with the Nixon Tapes
would know that reducing noise also entails the loss or distortion of the human
vocal range. In the case of the Nixon Tapes, the source audio was recorded at a
low gain and there little differentiation between frequencies of noise and
speech, compounding the difficulties of filtering out unwanted noise while
keeping legitimate speech.
No matter how sophisticated (or expensive) noise reduction software is, it is
still dependent on the original source audio.
Fine-tuning noise reduction software also
takes time and resources that are better channeled into transcription and
review. When the National Archives staff was attempting noise reduction with its
analog version of the tapes in the early 1980s, they decided in most cases that
the original recording was preferable and that very little staff time should be
devoted to noise reduction. Similarly, when the Marine Corps digitized its oral
histories from the Vietnam War several years ago, their staff decided that their
editing software, while somewhat effective, was too time-consuming in its
application and so they almost always used the original recording.
one long-time technical expert with expertise in expensive "forensic
audio" systems at the National Archives regularly says, “The best noise
reduction software is between your two ears.” While noise reduction software
can assist with the initial transcription of tapes where there is a particularly
pronounced hiss or buzz, no competent transcriber would use such a crutch during
the review process. The use of noise reduction software would be self-defeating
at that stage, since the reviewer would not being checking the transcript
against the actual audio, but rather, a modified version of it.
disclaimer also omits the fact that the Office of the Historian has had access
to digital audio since 2002, implying instead that such this is a recent
innovation. For the record, less than 5 out of approximately 40 volumes relied
on analog audio, many of them having been compiled prior
to the creation of the Nixon Tapes Group.
if the audio itself is the “document,” why hasn’t the Historian’s Office
put audio clips online, as has been repeatedly recommended over the last decade?
is a silver lining to the dark cloud looming over Foggy Bottom: the tapes review
process was completely reorganized in the summer of 2007 to ensure that each
conversation currently in the production process—with the exception of a few
volumes that had already crossed the publication Rubicon—meets the high
standards that should be expected from the Foreign Relations series.
If readers have any doubts, we encourage them to use the digital audio to
compare the transcripts in American
Republics, 1969-1972 with those of the flawed Soviet
Union, 1971-1972 volume.
SUGGESTIONS BASED ON TRIED-AND-TRUE EXPERIENCES
cardinal rule: When in doubt, CONSULT THE ORIGINAL AUDIO.
transcripts from the presidential recordings are impossible because the
source audio is, itself, far from perfect.
reduction software, while sometimes useful during initial transcription, is
not a panacea. When dealing with single channel audio from an analog
original, noise reduction also entails sound reduction and distortion. After
2007, the Nixon Tapes Group made only limited use of such software, relying
almost entirely on the unadulterated digital audio.
properly executed, group efforts are always better than individuals working
alone since transcripts benefit from the use of more than one ear (to catch
“unclear” and errors) and head (to provide subject-area expertise). (The
potential downside is groupthink)
there is any disagreement on what is being said, safe editorial practice
dictates marking segments as “unclear” or “unintelligible.” This is
the procedure NARA used when it was required to produce transcripts prior to
the general release of the recordings.
good equipment. Since the average office is actually a noisy location,
invest in some good noise canceling headsets, which block room noise without
altering the original audio.
reviews are a must. The revised system at the Department of State tracked
every conversation in every volume to ensure that each was reviewed a
minimum of 5 times by at least 3
different staff members, including one subject area specialist. This is a
standard that has yet to be eclipsed by any other institution in the world
that works with Nixon Tapes.
conversations, historically significant conversations, or those with audio
problems require yet more review than regular transcripts.
breaks every 20-30 minutes. Listening for hours straight will lower
productivity and degrade accuracy.
blast the volume. This hastens tone deafness, not to mention actual
what you hear, and what you read, especially when you are confronted with
nonsensical statements. (See summaries of errors below: 670-13
is a natural inclination to “hear” what you “see” on a page because
people are suggestible. Learn to avoid this tendency.
technology. Using a computer with digital audio, it is possible to fast
forward and rewind hours of tape almost instantly. It is possible to isolate
and repeat small sections. Electronic finding aids are easily searchable,
and spreadsheets with time codes can help an experienced researcher
immediately identify where to look for the relevant audio, especially when
there are close to 3,700 hours of Nixon Tapes.
reviewing a transcript, after completing the initial transcript, set it
aside and review it again the next day.
a corollary to embracing technology, use the best quality audio you can
acquire. Transcribing off digital audio is light years ahead of using analog
audio. Although the actual difference is only 10-15% compared to analog, a
good deal of tape work takes place on the margins.
a cost-benefit analysis. Try to keep things in perspective, balancing
resources. Is it really worth spending an hour trying to enhance 10 minutes
of audio? Remember, tapes are often repetitious, so unless you are dealing
with the only record of a significant event or a “Smoking Gun” type of
conversation, it might be worth tracking down another conversation with
similar content from the same time period. Similarly, does a conversation
need another review, or is listening to it 7 times over really enough?
doing a final review, go off the unaltered audio. An accurate transcript
should be able to be verified without noise reduction software or any such
common sense. Ask yourself if a troublesome passage makes any sense. There
are frequent non-sequiturs in transcripts. The recordings are what linguists
term “natural conversations,” so in many cases, the participants do not
speak in complete sentences. Some people stutter and have verbal missteps.
Others mumble or have accents. Frequently there is shared knowledge and
speakers use verbal shorthand that can make the meaning opaque.
Nevertheless, if something just does not seem right, be sure to check it as
many times as necessary to confirm or mark the segment as unclear. Better
ten “unclears” than one “cuddling.”
is key. Knowledge of events, geography, etc. is invaluable. One Department
of States historian managed to render invaluable help to the Nixon Tapes in
spite of the fact that he wears hearing aids because of his unmatched
subject-area knowledge. Inexperienced transcribers should not hesitate to
ask for guidance from their more experienced colleagues in dealing with the
plethora of verbal tics found in the Nixon Tapes (such as Nixon’s familiar
refrain, “and so forth, and so on”).
your ego at the door. We all make
reiterate the cardinal rule: When in doubt, check the unadulterated
audio—then check it again.
THE FLAWED TRANSCRIPTS
Relations of the United States, 1969-1976, Volume XIV, Soviet Union, October
1971-May 1972 (Washington, D.C.: U.S. GPO, December 2006) |
transcripts prepared under the pre-2007 quality control system. Reviewed twice,
possibly three times by two transcribers/compilers before being published.
Repetition and stuttering have been omitted. Note the number and degree
of substantive errors in these “interpretations.”
**Note that the corrected versions have been brought up roughly to the level of
the post-August 2007 quality control standards, although we did not have access
to the same resources as the State Department**
PDFs include time codes that correspond to the transcribed portions]
recommend right-clicking on the “Complete Conversation Audio” and
downloading the file to your computer for an easier listening experience]
Office Conversation No. 670-13, February 14, 1972, 1:04 – 2:25 p.m.
to note that Haldeman was not only present but participated in a number of the
yet numerous omissions (“I mean,” And,” etc.—also brief exchanges),
grammatical & editorial errors, et cetera.
to omission of Haldeman from introduction of editorial note, in at least one
instance, Haldeman was misidentified as Nixon.
(FRUS): And there will be another one in San Diego.
(Probably): And, then we had another one in San Diego.
(FRUS): —with the hope that we will lay off our preemptive air strikes.
(Probable): Maybe’s he’s invited you for lunch with the hope that we will
lay off of our preemptive airstrikes.
(FRUS): They think you are getting ready to club the North Vietnamese.
(Probable): They think you’re getting ready to clobber North Vietnam.
(FRUS): That is—
(Probable): That is a big deal.
second thing it made me think of was that—
(Probable): The second thing it may be, Henry, but the second thing they make be
thinking of is that they’re inviting—
can’t believe that they would tell you on the other side of the coin, now I
might be wrong, but they would have you for a private meeting and then proceed
to kick the hell out of us.
(Probable): I can’t believe that they would tell you, on the other side of the
coin—and I might be wrong—that they would have you for a private meeting and
then proceed to kick the hell out of us prior to that time.
(FRUS): Because that’s why [unclear].
(Probable): Because that’s why I need you to put the condition—stick with
(FRUS): If you accepted the meeting and then they kicked the hell out of us and
then we cancelled we’re in a [unclear] if you warn them in advance. Right?
(Probable): If you accepted the meeting, and then they kicked the hell out of
us, and then we cancelled, we’re in a bad way. We should warn them in advance
(FRUS): Because if they hadn’t they would have had people there looking at
(FRUS): Yeah. That they didn’t amount to anything?
(Probable): Because, if they hadn’t, they would have had people there looking
at the holes.
—and showing that they didn’t amount to anything.
(FRUS): And we’ve had another report that has been particularly—they
inflicted enormous casualties on some troop barracks.
(Probable): And we’ve had another report that in Vinh, particularly, they
inflicted enormous casualties on some troop barracks.
(FRUS): On the other hand, you and I know that you were going to go for broke
against the North.
(Probable): On the other hand, you and I know that you weren’t going to go for
broke against the North.
(FRUS): And then afterwards admitted Bhutto let you down.
(Probable): Then, afterwards, admitted to Bhutto that, “We [the Chinese] let
(FRUS): What they do is they’re asking for, cuddling for, the things we are
going to do anyway. Like troop withdrawal.
(Probable): What they do is they’re asking toughly for the things they know
we’re going to do anyway, like troop withdrawals.
(FRUS): So that the North Vietnamese will not forgive.
(Probable): So that, the North Vietnamese will never forgive them.
(FRUS): They’ve already objected in October so they—
(Probable): They’ve already rejected it in October, so that—
Office Conversation No. 720-4, May 5, 1972, 8:55 – 10:09 a.m.
(FRUS) Nixon: I want you to be rather cool, particularly outgoing with Dobrynin.
Nixon: I want you to be rather than cool, particularly outgoing with Dobrynin.
Nixon: They might [unclear].
Nixon: That’s the thing I said the other day.
Nixon: So that would be ineffective.
Nixon: And so, in net, it would be ineffective.
Nixon: Well then, that perhaps is the mess we’re in because we can’t bomb
unless we bomb now. We can’t bomb and then have—you can’t bomb and then
have them kicking us around while we’re in Moscow. You see, that’s the point
Thieu made which is tremendously compelling.
Nixon: Well then, that perhaps is the convincing reason, because we can’t bomb
unless we bomb enough. We can’t bomb and then have—you can’t bomb and then
have them kicking us around while we’re in Moscow. You see? That’s the point
that you made which is tremendously compelling.
Kissinger: One is does the United States put a Communist government into power
and allow itself and its enemies to defeat its friends?
Kissinger: One is: Does the United States put a Communist government into power
and ally itself with its enemies to defeat its friends?
Nixon: You see—look, Henry, there’s nobody that’s more aware, because I,
like you, one of the reasons we’re both in here, is that we both take a long
view, which goddamn few Americans do. That’s why I said that we put out a
little game plan if we wanted to cancel the summit first and then going after
them, which I think we’re absolutely right in not doing that.
Now that is something—
That’s good advice, because it’s something I’ve seen. I led you into
that—I led you out of that, yes I did.
Nixon: They’ve had it before. You see? Look, Henry, there’s nobody that’s
more aware, because I, like you—I think one of the reasons [unclear: we’re
both in here] is that we both take a long view, which goddamn few Americans do.
That’s why I said, “Henry, let’s put out a little game plan [unclear]
cancelling the summit first,” and they’re doing that, which I think we’re
absolutely right in not doing—
No, that is certainly not—
That’s good advice on the part of Connally.
That is certainly—
He has seen something I had not seen. And I led you into that.
I led you into that. Yes, I did.
Nixon: Now, here, the blockade plus, you understand—
Nixon: Now, here, the blockade plus the bombing, you understand?
Nixon: Well, everybody knows then, that I’ve thrown down the goddamn gauntlet,
and there it is. Do you want to pick it up? And, you see, I’m going to lift
the blockade as I’ve said. It’s not over yet—the bombing’s not over yet.
Nixon: Well, then everybody knows then that I’ve thrown down the goddamn
gauntlet, and there it is. And [do] they want to pick it up? And, you see, that
I’m going to live with the blockade, as I’ve said. Well, it’s an
Bombing is not an ultimatum.
Nixon: Even if it all goes down the tubes, we will be remembered as the ones who
went to China. And in the future, that’ll work out.
Nixon: You have to remember, even if it all goes down the tubes, we will
just—we will be remembered, as Clare Booth Luce says, as the ones who went to
China. And in the future, that’ll work out.
Nixon: Who could help us to do—all right?
Nixon: Who could help? Who else could do it?
Nixon: […] I saw the inevitability of McGovern, or Humphrey, or the only other
possibility is Teddy, who might be the worst of the three.
Certainly. No, McGovern’s the worst.
But anyway, as I saw that—McGovern would be the worst of the three for sure
Nixon: […] I saw the inevitability of McGovern or Humphrey, or, if they’d
have him, the only other possibility is Teddy, who might be the worst of the
Certainly the worst.
But any of them—
—in any event , because I saw that—No, McGovern would be the worst of the
three for sure if he gets in […].
Nixon: All right, I’m considering going […]
Nixon: All right, I have considered it all […]
Nixon: […] I assured Rogers and Laird, [unclear] let’s make another offer,
and have we agreed to offer this, and well, I don’t know if we have, and
they’re wining and bitching about it. Well, Henry, you know and I know this is
Nixon: […] I assured Rogers and Laird with regard to this. They said, “Oh,
let’s make another offer, and have we agreed to offer this,” I don’t know
whether we have, you know, and they’re wining and bitching about it. Well,
Henry, you know and I know that that’s just not true anymore.
Nixon: Oh, I hope they know, the guy across from me helped to break them
off—did you get that across?
Nixon: Well, I hope they know—it got across that they helped to break them
off. Did Porter make the case—?
WITH REVISED PROCESS
Relations of the United States, 1969-1976, Volume E-10, Documents on American
Republics, 1969-1972 (Washington, D.C.: U.S. GPO, July 2009) |
transcripts prepared and reviewed under overhauled, post-August 2007, quality
at least 5 times by at least 3 different transcribers/compilers.
has largely been omitted, although there is some repetition per compiler’s
-Note how accurate the transcripts are and how poor the audio quality is.
-Also note how the accuracy is consistent and the absence of major substantive
mistakes in these “interpretations.”
include time codes that correspond to the transcribed portions]
recommend right-clicking on the “Complete Conversation Audio” and
downloading the file to your computer for an easier listening experience]
Office Conversation No. 462-5, March 5, 1971, 8:30 – 10:15 a.m. Link
Office Conversation No. 523-4, June 16, 1971, Unknown Time between 3:41 and 4:30
Office Conversation No. 581-4, September 30, 1971, 10:07 – 11:03 a.m. Link
Office Conversation No. 735-1, June 16, 1972, 10:31 a.m. – 12:10 p.m. Link