Last January, I was sitting in the former headquarters of the Central Committee of the Communist Party of the Soviet Union, reading top-secret Soviet files about the Vietnam war. While turning the pages of a file, I unexpectedly came upon a startling document by a Vietnamese Communist general. It was a Russian translation of a report dated September 15, 1972 by Lieutenant General Tran Van Quang to the Vietnamese Worker's (Communist) Party politburo, detailing the number of American prisoners of war held by the North Vietnamese on that date. The document was the product of the former Soviet Armed Forces Main Intelligence Directorate (Glavnoye Razvedivatelnoye Upravleniye--GRU). Though I was unfamiliar with the details of the controversy surrounding missing American servicemen, I knew enough to realize immediately that the number of prisoners cited by the Vietnamese general--1205--exceeded the number of American POWs who were actually released six months later in early 1973, under the terms of the Paris Peace Agreement, by more than seven hundred. If the information in the document was accurate, its implications were likely to be explosive.

In the document General Quang described the American prisoners as being divided into three political categories--"progressives," "neutral," and "reactionary." The "progressives" would be released first. More important, Quang stated that Hanoi had created a separate secret camp system unknown to other prisoners. He acknowledged that in public Hanoi had deliberately understated the number of prisoners it was holding. Quang explained that Vietnamese communist policy was eventually to use the secret prisoners to achieve all its political, military, and economic objectives.

The archive in which I discovered the document bears the unassuming title of the Center for the Preservation of Contemporary Documents. It is just one of several major archives of the former Soviet Union's ruling party-state command center, which were requisitioned and sealed by Boris Yeltsin in the immediate aftermath of the abortive coup d'état of August 1991. I was engaged in research on projects that had nothing specifically to do with American prisoners of war and missing in action. The MIA issue was not an area of my professional interest.

My initial archival work in Moscow began in October 1992, when I worked on my own and was restricted to the pre-1953 Central Committee Archive. But by November I was affiliated with the Woodrow Wilson International Center for Scholars' Cold War International History Project, and enjoyed the special access to the post-1953 archives allowed by the Russians to such scholars.

The procedures for work at the archive are of some importance. Although the archivists knew of my topics, they did not know which files I would be requesting until the day I actually made the specific request. In fact, my own decisions about what files to request were made on a day-by-day basis and depended upon my ongoing reading of the Opisi (the folio catalogues). Following a request for files and their location by staff, the files would be screened by a senior archivist named Yuri Konstantinovich Maalov, who had the right to reject any particular request. As was the case with other researchers, all the files that I requested to see had not yet been officially declassified, and almost all were marked secret or top secret. Maalov refused me access to only two or three of all the files that I asked for during the time I worked there. But he did threaten to stop my archival access completely on other occasions. For anyone familiar with procedures at the archive, the suggestion made by Hanoi and some of its Western acolytes that the document was planted is absurd.

The first weeks of my work were devoted to the origins of Vietnam's decision to invade Cambodia. Then in mid-December I switched to a second project on the history of American intervention in Vietnam. As I knew that I might not have the time to research the entire historical period, and as the years 1972 and 1973 were central to one of the main arguments of my book, I began by asking for material for those years. All of the files requested came from the central committee department dealing with "Communist and Workers' Parties in Socialist Countries," i.e. ruling Communist parties. Most of them originated in the Soviet foreign ministry, and most consisted of reports from the Soviet embassy in Hanoi. A very few were KGB field reports or analyses, and a few others were reports of the Soviet Army General Staff, including those of the GRU.

On December 14, 1993 I requested ten files dealing with the year 1972. The set of requests was vetted and approved by Maalov on December 15 and I received them in the Archive reading hall on December 16. Because I was backed up with reading previously ordered files, and with Christmas and the New Year intervening, I did not reach the 1972 Soviet military file until January 1993. I did not discover any information about American POWs until I read the last section of the very last document in the file, on or about January 8. My initial reaction to what it contained was a mixture of shock and excitement. But it was only several days later, after telephone calls to better informed acquaintances in the United States, that I was certain that the document was of great significance. Only after several more days of inquiry at the U.S. embassy, did I realize that the U.S. government knew nothing about it.

Context and Evaluation

The document in question deals with more than the issue of American POWs being held in North Vietnam. That subject is addressed in only eight pages towards the end of its twenty-five pages. The rest of the document deals with two other issues: the attempt of the Vietnamese Communists to seduce disgruntled South Vietnamese generals and politicians into joining a coalition government with the Communists; and the dispatch of a team of assassins to carry out a campaign of terrorist subversion in South Vietnam.

The document was located in a file of the Soviet general staff which contained other documents all dated and pertaining to events occurring during the year 1972. In contrast with the files containing documents of the Soviet foreign ministry for 1972, the Soviet military files contained information of a highly sensitive nature and indicate that ties between the Soviet and Vietnamese military were close and informal at this time. This is not surprising, as between 1968 and 1972 the Vietnamese Communists shifted from neutrality in the Sino-Soviet dispute to a pro-Soviet foreign policy. Hanoi's army was re-equipped from the Soviet arsenal and reoriented itself in line with Soviet military doctrine. More Vietnamese officers were trained in Soviet military academies, and closer relations evolved between Soviet military advisers and Vietnamese military officers.

Clearly the document is the product of Soviet intelligence work. Having read the entire Soviet military file, I conclude that the Soviet military's GRU had a Vietnamese agent placed at a very high level of the Vietnamese party. This conclusion is critical for evaluating the document at issue. For it entails, among other things, that Quang's speech was prepared for the Vietnamese politburo, not for a Soviet audience.

Furthermore, the style of the document is clearly that of an oral report, not a written speech. This is most apparent in the section in which Quang carefully recites statistical data with no concern for sentence structure. It is therefore likely that the document is the translation of a transcript (either written notes or a tape recording) of a speech by General Quang. It is also likely that the translation was undertaken by a GRU case officer from the notes or tape recording provided by a well-placed Vietnamese agent. This assumption makes it easier to explain the document's possible factual inaccuracies.

More important is the endorsement of the document's authenticity and value by two key Soviet officials at the time. The then head of the Soviet GRU, General Pyotr Ivashutin, signed an executive summary of the document which was sent to the central committee. The member of the CPSU secretariat responsible for relations with ruling Communist parties, Konstantin Katushev, wrote a handwritten instruction across the front of Ivashutin's executive summary ordering his deputy to prepare a brief report on the prisoners of war for the CPSU politburo. These officials clearly believed that the report was significant and basically accurate. Ivashutin had by that time held his important post for twelve years, and his judgment must carry considerable weight. Both he and Katushev were well schooled in the art of political disinformation and deception, and were most unlikely to have been fooled by any effort in that direction on the part of the Vietnamese. Further, if as I have suggested this document, like others in the file, was passed by a Vietnamese agent, the question arises: What motive could such an agent have had for deceiving the Soviets on such a matter?

These considerations lead to the conclusion that the original Vietnamese source of the Russian document was probably an authentic Vietnamese transcript (handwritten notes or tape recording) of an oral report by General Quang to his politburo. But the Vietnamese transcript was then translated into Russian, with some editorial comments by the Russians, a process that allowed for error. Further, the translation may have had to be undertaken in a rushed manner, with the transcript available for only a limited time, which would have increased the chance of errors.

A fundamental starting point for any evaluation of the Soviet GRU document is to recognize that the Pentagon's database on Vietnam POWs/MIAs rests heavily upon information supplied by those former prisoners who returned to the United States in 1973 during Operation Homecoming. If, as the GRU document states, there was a separate prison system, of which the returnees knew nothing--and keeping them in ignorance would have been precisely the point of holding people separately--then the Pentagon database cannot be used for a conclusive evaluation of the GRU document. This is true not only for quantitative data (numbers of prisoners, numbers of each rank) but also for qualitative data (the mixing of ranks, special training, and other characteristics of the prisoners). The frequent refrain of Pentagon officials--that "nobody who came back reported this"--is therefore simply beside the point. Furthermore, the challenge to the quality of Pentagon intelligence by Russian sources must be taken seriously. After all, the Soviet military intelligence was politically and physically much closer to the Vietnamese Communists in 1972 than was its American equivalent.

Does this mean that the Russian document is completely accurate? Not at all. General Quang may have been misled by his subordinates on many aspects of the POWs. Some of the names in the document were garbled. The ranks and level of training of the prisoners may have been inflated, which could reflect incompetence or attempts by lower level commanders and interrogators to impress superiors. Again, prisoners may have misinformed interrogators in order to secure better treatment for themselves. And there is the previously discussed problem of a hurried translation.

Nevertheless some key factual aspects of the document have already been confirmed. As the document predicted would happen, three prisoners were released within days. The document's claim that prisoners were placed in one of three political categories was reported long ago in Admiral Stockdale's prison memoir In Love and War. This political categorization did not entail physical separation, though some, not all, in the "reactionary" category may have been kept in a separate camp system. We know that some of those who returned in 1973 had been held separately, unknown to the other returnees, until the very last moment. This confirms the possibility of a separate camp system holding hundreds of others. And the assertion that some prisoners had undertaken cosmonaut training is confirmed by the case of now Admiral Robert Harper Shumaker, a Navy flier shot down in 1965 and returned in 1973.

More important is the correlation of Hanoi's negotiating stance, as enunciated by Quang, with the secret negotiating position of the Vietnamese Communists at the time. Recently declassified cables, and the testimony of Dr. Henry Kissinger, confirm this.

There are three factual claims made by the document which are matters of contention: Was Quang a lieutenant general in 1972, as the document claims, or was he at that time only major general, as the Vietnamese Communists claim? Did Quang hold the position of deputy chief of staff, as the document claims, or was he serving in that capacity only during 1959-60 and from 1974 on, as the Vietnamese Communists subsequently claimed? And was a 23rd plenum of the Vietnamese politburo held in 1972 as the document claims, or was it held in 1974 as the Vietnamese Communists have subsequently claimed? These are matters about which the independent evidence is mixed.

The former U.S. Embassy Saigon's 1972 report on the membership of the Central Military Committee lists Quang as deputy chief of staff, while a biographical report from Vietnam dated November 1992 lists him as holding a field command but not the deputy chief of staff position. Yet he could have held both positions. There were always several deputy chiefs of staff, and some of these also held field commands. The failure of the 1992 North Vietnamese publication to list Quang in a deputy chief's position may simply reflect the post's lesser significance relative to the field command.

As for Quang's rank of lieutenant general, this was what he was described as in another Russian document, purportedly based upon a June 26, 1972 report by Quang to the politburo and having nothing to do with the POW issue. This suggests that if there was an error, it was a genuine error by the Soviet case officer or Vietnamese agent, and had nothing to do with either the issue of prisoners or an attempt to deceive.

Finally the question of whether there was a 23rd politburo plenum in 1972 or in 1974 as the Vietnamese subsequently claimed is also a minor issue, and arises from a reference in a section of the document dealing with issues unrelated to the American POWs. There was a 21st plenum in 1972, and either the Vietnamese agent or the Soviet case officer may have misheard the word "21st" for "23rd."

The only incontrovertible and strange error in the Russian document--the conflation of a General Ngo Dzu with politician Truong Dinh Dzu into a nonexistent "General Ngo Dinh Dzu"--is minor. It does not appear in the section of the document dealing with the American POWs. Read carefully, Quang seems to be speaking about the politician first and the general later. The error, likely made by the Soviet GRU translator, more familiar with generals than civilians, is perfectly explicable if we assume an audio tape source.

We must distinguish between what data was primary and what was secondary in the report. The most basic facts are that a senior North Vietnamese general told his politburo that they were holding hundreds more American prisoners than the number they had told the world about; and that the Hanoi politburo planned to use these prisoners as a device to extract military, political and economic concessions from the United States. These are matters of clear meaning which are not easily subject to linguistic ambiguity or error.

This central aspect of the document is given contextual support by an examination of the entire file in which it was located. It is very clear from reading all of the Soviet military intelligence reports emanating from Vietnam in 1972 that the Easter Offensive of that year was a serious military failure for North Vietnam. Hanoi had become pessimistic about its ability to win militarily. That is why holding back prisoners was important--the communists needed the American prisoners as a bargaining chip.

Hanoi almost certainly held back hundreds of prisoners in 1973. The most controversial aspect of the document, Quang's claim to be holding 1205 Americans, has been independently corroborated.. In 1971 a North Vietnamese military doctor, Dang Tan, who defected to the South, told the U.S. government that North Vietnam was holding hundreds more prisoners than they had acknowledged. In 1979 a Vietnamese communist defector named Le Dinh told the Defense Intelligence Agency (DIA) that in the mid-1970s his colleagues had spoken of holding seven hundred American prisoners to be used as a "strategic bargaining asset."

Why, then, has this card never been played subsequently? Perhaps all the prisoners were executed in 1975 or some time later, after the unexpected collapse of South Vietnam seemed to make such a bargaining asset unnecessary. It is at least possible that some prisoners are still being held as the very last card, in case the attempt to gain normalization through investment offers to American businesses does not sway Washington.

Back in Washington

An objective evaluation of the accuracy of the information in Soviet GRU document is of fundamental importance for the formulation of American policy towards Vietnam. The view that the document must be taken seriously has been endorsed by Henry Kissinger, Zbigniew Brzezinski, Richard Pipes, Adam Ulam, and Samuel Huntington. All are distinguished scholars, with either extensive knowledge of Soviet affairs, or extensive experience in dealing with Moscow, or both. Although nobody of their intellectual caliber has taken an opposing view, from the very beginning the Clinton administration has not handled the issue issue with the appropriate seriousness.

My first inclination after discovering the document in January was to give a copy of it immediately to the New York Times. On reflection I quickly concluded that as some prisoners might still be alive and their lives at risk, I was morally obliged first to inform the United States government. To this end, immediately upon my return to the United States on January 31 I contacted a Congressional staffer whom I knew well, and sought his assistance. He faxed a letter to the chief of staff of the NSC, Nancy Soderberg, on February 1, and followed up with a phone call to her on February 3. But in spite of this staffer's high personal reputation, no one from the NSC contacted me until after I had, more than one week later and in frustration, brought this matter to the attention of the office of the secretary of defense through the personal efforts of Professors Richard Pipes and Samuel Huntington of Harvard.

On February 9 I was finally contacted by Deputy NSC Adviser Samuel "Sandy" Berger who agreed to meet me. He asked if I would give him a copy of the document. I suggested that we discuss it. At our meeting on February 11 I explained to Berger that I did not want the document to reach the bureaucracy as I feared that it would either be leaked or treated in a negatively biased way, because it ran counter to everything the relevant bureaucrats had been saying on this matter for over twenty years. But I provided Berger with the names of three "authenticators" of the document--Zbigniew Brzezinski, Richard Pipes, and Mark Kramer, all Russian-literate scholars who were familiar with Soviet documents in general and who had read this one in particular. (The first was also a former national security adviser to the president, the second a former member of the NSC staff, and the third a young scholar whom NSC chief of staff Soderberg knew personally and had consulted many times). Moreover, I emphasized that this document was the beginning of the story, not the end. The fate of these men could only be determined through other Soviet documents, located in other Russian archives to which I currently had no access. I said that I wanted to pursue this story in Moscow with the U.S. government's help.

Berger did not ask for the document, obviously aware of my reluctance to hand it over, although in fact I had come prepared to show it and my English translation of the key sections to him. Instead he asked for key pieces of information from it so that it could be checked out by the relevant government experts. I provided him with a prepared two page memorandum which provided this, and which also explained where the document came from, why it was credible, and how the matter could be pursued further in Moscow.

I requested that my memorandum be kept out of the hands of the bureaucracy. Although I had never met nor even heard of the Department of Defense's MIA bureaucracy before that week, I instinctively feared that their concern with this issue would amount to nothing more elevated than defending personal reputation and political damage control. Although Berger said he would contact me again in a week, that was the last I saw or heard of him for five weeks, despite several follow-up calls on my part.

Meanwhile Zbigniew Brzezinski, to whom I had shown the document on February 10, informed Tony Lake of the document's contents and political credibility over lunch on February 18. According to Brzezinski, Lake was interested and shocked by his account. But this intervention produced no discernible effect.

NSC chief of staff Soderberg sent the detailed memorandum which she had received from the congressional staffer, not the briefer memorandum I had given to Berger, to the Department of Defense for evaluation. Secretary of Defense Les Aspin, who had been informed by Pipes of the seriousness of the document and my willingness to publish it, instructed his subordinates to try to get a copy of the document from me. Deputy Assistant Secretary of Defense for POW/MIA Affairs Ed Ross and Colonel Stuart Herrington of Task Force Russia (the Pentagon's support group for the U.S./Russia Commission on POW/MIAs) acted on that instruction, but were unsuccessful. At some point somebody decided that the U.S. government would try to acquire its own copy through the Commission.

But in the interim the analytical division of the Defense Intelligence Agency and Ed Ross had already made up their minds about the value of the document. When Sandy Berger finally called me back on March 17th, he asked me whether the document referred only to Americans. He told me that he had been advised that if the 1205 figure referred to other nationalities as well as Americans, then there was no problem for the DOD data base, but that if the document referred only to Americans then it was inconsistent with the information that the government had. I expressed to Berger my absolute certainty that the document was discussing only Americans.

In subsequent weeks I had conversations with the belatedly appointed NSC specialist on Asia, Kent Weiderman. The career foreign service officer seemed genuinely concerned about the document and indicated that he took its contents seriously. Weiderman said he would try to arrange a meeting for me with Lake and Berger. He asked whether I would be willing to bring a copy of the document and I agreed to so do. But he was never able to organize a meeting.

It is hard to avoid the conclusion that either resolving the MIA question was a low priority for the newly-installed National Security Council, or else the NSC did not believe that the Soviet document was an accurate or valuable source--or both. I certainly believe the second to have been the case. For the NSC had done what I had naively expected them not to do: called upon and accepted the expertise of the permanent bureaucracy--which had a record to protect--before making any other effort to pursue the matter in other ways.

On April 8 the head of the American side of the U.S./Russia Commission on POW/MIAs, Ambassador Malcolm Toon received his copy of the Soviet GRU document from the Soviet joint head, General Volkogonov. I arrived in Moscow to continue my academic research on April 9th. On April 10 the Russian daily newspaper Izvestiya published a report of the handover of documents, with extracts of key facts from the GRU document on American POWs. On the evening of April 11th, realizing that the leaking of the document in the U.S. could only be days away, I took a taxi to the New York Times bureau in Moscow, and gave a copy of the document to the journalist Celestine Bohlen. The next morning I flew back from Moscow to the United States, to do what I could to explain the document and its background to the American people.

The Bureaucracy and General Vessey

The official defense department view of the Soviet GRU document (labeled by the DOD as the "1205 Document") was presented in the DOD POW/MIA Newsletter of July 1993. It is worth quoting the key passage summarizing their evaluation:

The "1205 document" appears most credible in its first section, about political operations planned for South Vietnam. The report, however, also contains numerous errors and inconsistencies, particularly on the section on POWs. While portions of the document are plausible, evidence in support of its claims to be an accurate summary of the POW situation in 1972 are far outweighed by errors, omissions, and propaganda that detract from its credibility. As additional information becomes available, the Department of Defense and other U.S. government agencies will continue to assess the document. At this point our bottom line judgment is that the document and the information contained in it suggesting the Vietnamese held more than 600 POWs is not accurate.

What are the "errors, omissions and propaganda that detract from" the credibility of the Soviet GRU document? The DOD did not feel it worthwhile to cite any of them. Presumably the American people, including MIA family members, were expected to take the DOD's evaluation on trust. Somewhat bizarrely, the DOD evaluation is still officially "ongoing," in spite of conclusions being already reached. But let us examine what available written and verbal statements by DOD members, the public statements by General Vessey (apparently based upon briefings by DOD and State Department officials), and the State Department intelligence unit's report show us about how they evaluated the evidence.

On April 12, after publication of key extracts of the document in the New York Times, the director of the Defense Intelligence Agency's (DIA) special office on prisoners of war and missing in action, Robert Sheetz, wrote a memorandum to the acting deputy assistant secretary of defense for POW/MIA affairs, Ed Ross. The memorandum stated: "DIA believes that this document is referring to both U.S. POWs and to allied POWs, particularly ARVN commandos. The confusion probably lies in an inaccurate translation from Vietnamese to Russian."

Sheetz's analysis is preposterous. Anybody remotely familiar with published and secret Vietnamese Communist documents knows that Hanoi always distinguished allied from American troops, with the former described as "puppets." The use of such distinctive terms as "American" (My in Vietnamese, amerikanyets in Russian) and "puppet" (nguy in Vietnamese, marionyetka in Russian) is so frequent in communist writings as to make their confusion by any competent translator inconceivable. In an earlier report to the Vietnamese politburo dated June 26, 1972, Quang drew the distinction frequently. Clearly, his use of the term "1205 amerikantsi" in September could refer only to 1205 Americans.

For the month before the Russian document became public, and before the DIA had actually seen the document, this had been the DIA's and the DOD's unofficial interpretation. It continued to be such, through the medium of off-the-record press briefings, for another week, until my article in the Washington Post (April 18, 1993) noting Quang's earlier speech and its terminology.

On April 18, General John Vessey led a mission to Hanoi on behalf of President Clinton, to pursue the POW/MIA issue. The materials received by the Vessey mission--documents suddenly "discovered" by Hanoi officials after the revelation from Moscow--did not address the charges in the Russian document, because they do not discuss the fate of servicemen still unaccounted for. And because they are accounting records of the known prison system, the Vessey documents neither confirm nor falsify the Russian document's assertion of a separate prison system. Yet Vessey misled President Clinton by telling him that they challenged the Russian document.

More disturbing were Vessey's public statements. In Hanoi he said, "I have no reason to disbelieve" General Quang's denial that he had ever made the report contained in the Russian document. But what did Vessey expect Quang to say? After all, if the Russian document is accurate, General Quang is complicit in a huge crime. Even if General Vessey had possessed no means of judging General Quang's credibility, common sense should have suggested that any criminal suspect's mere denial of guilt is not sufficient ground for suggesting innocence. But Quang and his alleged accomplices do have a record by which we can evaluate his denial. He has been a senior figure in a regime which has slaughtered tens of thousands of its own citizens in peacetime as well as war. He has served a regime which for twenty years denied that it was controlling, supplying, and eventually directly participating in the war in South Vietnam. He has served a regime which for over ten years denied that it had military forces stationed in the independent nations of Laos and Cambodia. A candid listener might have accurately commented "I have no reason to believe" General Quang's denial--or, if prudence required it, remained silent.

Equally disturbing was General Vessey's attempts at textual analysis. Though this was not his assignment, he undertook the task anyway. His evident purpose was to show that the document was not accurate. Yet although he had had ten days to examine it before speaking publicly, he still managed to misrepresent its contents. On the day after his return from Hanoi, Vessey stated that the document claims that American prisoners were segregated according to their political views, which is inconsistent with what we know from returnees. Not so. The document suggests that Hanoi divided the prisoners up analytically, not physically.

Vessey stated that the document wrongly claims that prisoners were released according to their political views, when in fact release in 1973 was in accordance with date of capture. But Vessey is confused. Only the final process of release of those already selected was determined in accordance with date of capture, not the selection of who would be released. Moreover, since Vessey does not possess the Vietnamese list of who was in each category, how can he know whether or not the group of 591 finally released in 1973 did not include the 368 Hanoi deemed political "progressives"?

Vessey stated that the document claims that senior officers were segregated from others in the camps. But the document is highly ambiguous here. One passage might seem to be suggesting that. But the total number of senior officers Quang claims were held in four separate prisons is less than the total number of senior officers that he admits, elsewhere in the document, to be holding. Reading the whole document carefully, one could also conclude that 355 senior officers were segregated while others were mixed with lower ranks.

Next General Vessey points (correctly for once) to a claim by the document that after the Son Tay raid the prisoners were dispersed into several different camps. Not so, says Vessey; what happened after the Son Tay raid was that prisoners were concentrated in fewer camps, not dispersed. Yet the Pentagon's own evidence from returnees--contained in the DIA's report of the POW camp system, declassified by the Senate Select Committee last year--turns out to be neither of concentration nor of great dispersal.

Finally Vessey asserts that there could not have been 1,205 prisoners held because neither the returnees nor U.S.