The second basic failure has been -- until very recently -- the relative neglect by US intelligence of the wellsprings of the enemy's power, the Communist Party and its related collectively called the "infrastructure". In early 1967 for example, the CIA -- despite a plethora of evidence -- was scarcely aware of the existence of its principles adversary in Vietnam, the Viet Cong Security Service, which roughly equivalent to the Soviet KGB. Likewise, we are only now beginning to inspect the strength and effectiveness of the enemy's large and well-organized Proselyting apparatus, which is devoted to undermining the morale of the Allied armies, and which has thousands of contacts in ARVN Part of the reason for the neglect has been the omission by intelligence components in to study the infrastructure systematically. Even now, the number of intelligence officials in Washington with a working knowledge of the infrastructure is woefully inadequate.

Furthermore, the imbalance of the intelligence -- a condition by the probable presence in South Vietnamese government and of over 10,000 Viet Cong agents, compared to fewer than a hundred to have in -- has had several overlooked consequences. First, the has allowed the to deploy the forces far more efficiently than the and numbers, as pat are extremely misleading do the imbalance gives the enemy and therefore allows him to control his losses: an which question any policy which calls for a war of attrition. And the to be sell-reinforcing. The enemy has in GVN ranks, the for to agents among the Vies A may be that the that it has little hope of except in the long torn. The third been the frequent of the Allies to the most prominent example of this was our to and initial of the enemy's largest far, the Part of the reason for our failure to guess the enemy's intentions at Vet was that we had greatly his 14.3-67, agreed to by the entire and less than three months before the underestimated his numerical strength in the and his from the north. It effectively of an attack as large as the enemy launched. Also, the intelligence had not fully addressed the capability of the apparatus to infiltrate large numbers of soldiers into the cities. We thus hardly considered the possibility of a as the which occurred. The final reason for the failure to predict the Tet attack was the of in the enemy organization who could have told us what he was going to do. The CIA's relative look of in espionage had at three we started late (in mid-1966); few of field Vietnamese (about three); and CIA officials have only an imperfect knowledge of the enemy. The CIA case officer to the provinces prior to 1968 has more than two of on Viet cong and post. Three categories: A. A. for a of to examine the of intelligence in Vietnam and The is on the that the conflict has that has deficiencies, which to be and corrected. B. Short-term put forward to cover specific which in Vietnam. The most the which has greater for the Viet cong than any of our previous programs, either civilian or military. For one recommendation is designed to from degeneration into a like we gave in in the past. C. Longer term recommendations principally designed to correct basic to research, which has been at the root of in Vietnam. Until the are the US the risk of finding itself, in to one it was in in late 1964 to our Should such a arise, US intelligence mast be in a position to tell policy-makers what they are getting into, so that they can -- on the of careful intelligence -- the cost of policy alternatives. My recommendations are neither all- nor absolute. They are set out in the hope that intelligence will take a serious look at itself to determine how it can do a better job then the performance it has turned in during the struggle. We move on to Adams's discussion of CIA operations on the ground in Vietnam, as well as his recommendations. The Lack of Vietnamese Speakers 96. The number of CIA case officers in the field capable of speaking Vietnamese has always been tiny. the current number in Vietnam, I believe, is in the neighborhood of three. I doubt whether this figure has ever been greatly exceeded. 97. A number of excuses have been advanced for the phenomenon. They include arguments that: A. The Agency, which has had to supply unusually large numbers of officers to Vietnam, has not been able to invest the extra personnel to language training. B. Such a program would be expensive. C. It might not produce the desired results. D. Vietnamese interpreters can handle the problem, and, E. Agency of third-country interpreters -- of whom there are a few in Vietnam -- are sufficient for operational needs. 98. Although frequently voiced, the first three arguments can be dismissed out of hand. The fourth, that Vietnamese interpreters are adequate, overlooks considerable evidence that the Communists have concentrated a great deal of espionage talent at their recruitment. The last, that we have a stable of presumable reliable US and third-country interpreters, is the most substantial. It can be met by the usual reasons given as to the desirability of espionage case officers speaking the local language. I need not elaborate on these, but would merely provide what seems to me a good example of the advantages to be derived from fluency in Vietnamese. 99. The consistently best reporting on the Viet Cong is thought by many to be RAND's Dinh Tuong (DT) series of interviews of Viet Cong captives and soldiers. The person responsible for the interview was an American named David Elliot. He spoke fluent Vietnamese but seldom saw the prisoners and defectors. The high quality of his product was due primarily to the excellence of his Vietnamese interviewers, none of whom spoke English. Mr. Elliot was able to find such good interviewers because he was not limited by language to recruiting among English-speaking Vietnamese. All but one of his interviewers were elderly Vietnamese gentlemen, who rarely speak English, but who command the respect of the usually young Viet Cong prisoners and defectors. Mr. Elliot's reports were frequently superb. CIA province officers have yet to duplicate his performance. Paucity of Training on the Viet Cong 100. Until recently, case officers going to the field in Vietnam received little training on Viet Cong organization and techniques. Their training was restricted to such subjects as general espionage or interrogation techniques, Vietnamese history, and the organization of Allied programs. The amount of formal instruction on the Viet Cong seldom exceeded two hours. Once in Vietnam, officers going to the VC Branch were able to spend a number of weeks "reading in" on assorted material concerning the Viet Cong. Officers assigned to the provinces usually did not even have this opportunity. Thus the average CIA province officer arriving at his post was not only unable to speak Vietnamese, but was largely unaware of the nature of his target. Frequently, his reporting has reflected it. 101. In August 1968, at the request of the head of the DDP-run South Vietnamese Operations Course (SVNCC), I instituted a two day course on Vietnamese Communist organization and techniques for Agency personnel going to Vietnam. The two-day course has now been given on three occasions and was expanded to three days in January 1969. I have three comments: A. The course is by far the most detailed instruction given on VC. organization by any agency of the US government. B. It is superficial, hastily put together, and inadequate. C. The head of SVNCC, who had been trying to start up such instruction for some time, had been unable to find anyone willing to take on the task. Conclusion 102. The late start and the neglect of basic preparations have meant that the CIA has misspent valuable time and scarce espionage talent in operating against the Viet Cong. For example, one of the first major programs of the VCB was a simple case of mistargetting-through ignorance. 103. The name of the program was TUJOCKEY. Mounted by the VCB in the latter half of 1966 and continuing through 1967, its purpose was to split (among other endeavors) the Party and the National Liberation Front (NLF.) A routine familiarization with the relationship of the Party apparatus and the Committees of the NLF, particularly at the higher echelons (at which the program was directed) would have suggested its fatality. The Front is, of course, a wholly-owned subsidiary of the Party, with virtually no independence. The time, money, and personnel expended on TUJOCKEY, then the CIA's biggest espionage operation against the VC, would have been better spent elsewhere. The effort expended on TUJOCKEY was not wholly wasted, however, In on sense, it served as a training ground for recent operations. Certain of these appear to be relatively successful. The problem with them is that they are so late. -24- V The Atmosphere Within the Intelligence Community 105. As suggested in the foregoing paragraphs, a lack of foresight, a neglect of fundamentals, and an absence of clear central direction, have characterized the US intelligence effort in Vietnam. These are primarily technical problems. the most basic question is not technical, but atmospheric. The temper within the intelligence community during much of the war has not been conducive to honest appraisal. 106. The mood has had several distinct characteristics. First, there has been a frequent lack of courage in advancing ideas conceived of as unpleasant. Timidity and vacillation at the top have seeped to the lower ranks, so that many issues of real or potential moment have remained submerged among the underlings. The common reason advanced for such timidity has been "political pressure". Although political considerations cannot be avoided in conducting intelligence, the excuse is weak. Intelligence which lacks honesty lacks utility. 107. Second, the atmosphere has often been charged with a want of candor. Forthrightness has all too frequently given away to indirection, usually at the expense of clear English. Intelligence conferences over enemy numbers, for example, were elaborate bargaining sessions rather than a careful weighing of evidence. Middle men bartered in corridors, while the principles pored over clauses in the contract, designed more for the press than for policymakers. The end products until March 1968, were "agreements" which obscured enormous differences. 108. Third, there has been, until very recently, an avoidance of self-criticism. although the United States has been losing a major war against a minor power, criticism has been met with delay, evasion, and attempts to explain away past failures. The lack of critical introspection contrasts sharply with the practice of the Viet Cong, whose report writers are required to dwell on weakness. 109. Finally -- in large measure because of conditions already touched on -- jumble and confusion have often reigned. Considered reflection born of exhaustive study has been abandoned frequently for headlong rushes into complicate problems. All too often, the answers preferred by the intelligence community have reflected the manner in which they were sought. -25- VI Recommendations 110. My recommendations fall into three main categories: A. A general recommendation for a Board of Inquiry to examine the overall conduct of US intelligence in Vietnam and elsewhere. B. Short-term recommendations concerning Vietnam. C. Longer-term recommendations transcending Vietnam. 111. Certain portions of these recommendations have been put forward elsewhere by other people. Where so, my suggestions are made in order to add my voice to other. Recommendation for a Board of Inquiry 112. I respectfully recommend that the Executive Branch of the Government appoint a Board of Inquiry to investigate thoroughly the conduct of the US intelligence community in Vietnam and elsewhere. The threefold purpose of such an inquiry would be, first, to find out where US intelligence has failed in the last five years, particularly in Vietnam, second, to ascertain where shortcomings still lie, and third, to recommend measures to avoid similar deficiencies in the future. 113. I respectfully suggest that objectivity would best be served if the Board were headed by a person uninvolved in our policy in Vietnam and unconnected with any components of the US intelligence community. 114. I further recommend that the Board consider taking certain broad avenues of inquiry, to include: A. The direction, organization and management of intelligence research. B. The targeting and preparation of clandestine operations, including such matters as training and language policy. C. The overall control and coordination of military and civilian elements of the intelligence community. Short-term Recommendations Concerning Vietnam 115. How long and how heavy our involvement in Vietnam will be is far from clear. Given the uncertainty, the US intelligence community ought to prepare for the long haul. My recommendations are fundamental and relatively inexpensive. They are advanced below in broad outline. I will supply more detailed recommendations if requested. 116. Short-term Recommendation one: US intelligence should embark on a community-wide program to educate Allied officials more thoroughly on Viet Cong organization and techniques. the program should include: A. The creation of an inter-Agency committee to determine what training on the Viet Cong needs to be given, who is to give it, and who is to get it, both in Vietnam and the United States. B. The creation of a standard two-week course on Vietnamese Communist civilian and military bureaucracies. The course should be made available as soon as possible to all appropriate Allied intelligence and security officials, including Vietnamese, serving at district level and above in Vietnam, and to American intelligence officials working on Vietnam in the US, including researchers, desk officers, and training officials. If requested, I will supply a suggested course outline. C. The creation of specialized courses on specific parts of the infrastructure. For example, I would suggest the putting together of a one-week course on Viet Cong intelligence organizations (including the Cue Nghien Cuu, Military Intelligence, Military Proselyting, and the Security Service) for counter-intelligence officers going to Vietnam. D. The writing and maintenance of a series of basic handbooks on specific parts of the Viet Cong infrastructure for small libraries (say, 30-40 volumes) on the infrastructure to be maintained at district level and above. For example, a handbook on the enemy security apparatis has already been written, but needs updating. A handbook on the Military Proselyting organization has yet to be published. The programming of such handbooks should be determined centrally. The handbooks should be classified "For Official Use Only," or "Confidential", so that they may be given wide dissemination in Vietnam. Periodic inspections should be arranged to ensure that the libraries are kept up to date. 117. Short-term Recommendation Two: CIA Deputy Directorate of Intelligence should create a task force of at least one dozen researchers to conduct in-depth research on the enemy, particularly his party bureaucracies. Properly coordinated with the field and with the DDP, such research need not duplicate that done by the Station's Research and Analysis Branch, but ought to complement it. Most important, the task force would give Washington a capability it does not now possess: an ability to render coherent and detailed judgments on Party affairs. -27- 118. The task force should have at least three purposes: A. To monitor and evaluate in detail the PHOENIX program in order to measure its effect on the Viet Cong apparatus. The evaluation should include a continuing analysis of PHOENIX statistics, and, most important, qualitative judgments on the results of PHOENIX operations. B. To prepare substantive studies for policy-makers on certain basic but largely neglected subjects. These include: I. A detailed study of the efficiency and impact on the Allied war effort of Viet Cong intelligence and security organizations (to be done in cooperation with the CI Staff of the DDP.) II. A study of the impact on the Allied war effort of other VC covert action operations. (See Paragraph 67). III. Continuing studies on the policy and structure of various Viet Cong bureaucracies, particularly including the security service, and the military proselyting apparatus. C. To perform certain support functions for other CIA and community intelligence components. These could include the production of handbooks on the infrastructure, the preparation of interrogation questionnaires for various types of VC prisoners, the supply of instructors to train US officials going to Vietnam on VC organization, and related tasks which PHOENIX and other organizations might propose. 119. If created, the task force could either be given independent status within the DDI, or assigned to a specific DDI office. In no case should it be swallowed up by existing components, or put to such tasks as producing "current intelligence" on Vietnam, which already has a full division of the Office of Current Intelligence occupied. Provision should be made to allow it adequate space, including a library for storage of primary materials on the Viet Cong: -- for example, captured documents, POW interrogations, and defector reports. 120. Short-term Recommendation Three: The intelligence community should thoroughly reappraise the goals and operation of the PHOENIX program. The reappraisal should include: A. A meeting, as soon as possible, of appropriate components of the community, to devise a working definition of who belongs to the Communist infrastructure. I would recommend that the definition include a spectrum, which would distinguish infrastructure members by echelon, job description, and importance. -28- My own predilection in reworking a definition would be to allow for the inclusion in the "infrastructure" of many more Viet Cong than are presently taken into account in MACV and CIA working definitions. My view stems from the belief that many of the tasks performed by low-level personnel in the Communist structure are important, and damaging to the Allies. B. The creation by PHOENIX of a reporting procedure which would allow for a comparison of its "eliminations" to a measurable base, preferably one such as envisaged in Subparagraph A above. C. A retroactive inspection of PHOENIX's past reporting, to determine, as far as possible, the damage the program has inflicted. The retroactive look should include a careful appraisal of the quality of personnel eliminated, together with an estimate, if possible, of how many "neutralized" officials have rejoined the Viet Cong, and the extent to which the VC have been able to fill any voids created by PHOENIX. D. An assessment of the counter-intelligence problems the program presently faces, and a determination of what measures can be taken to meet them. E. An assessment of what the PHOENIX program can realistically expect to accomplish, within given periods of time. 121. Short-term Recommendation Four: The intelligence community should reassess personnel policies for officials going to Vietnam, with an eye to increasing professionalism and length of service there. Specifically, I should recommend that: A. The armed forces increase the length of tours of intelligence personnel from one year to at least eighteen months, or more where practicable. B. The CIA set up a program of incentives to persuade its officers in Vietnam to stay beyond their regular tours. I would suggest that consideration be given to paying appropriate personnel additional funds over and above their regular salary and allowances to persuade them to extend.* The cost would be tiny compared to overall Vietnam expenses. *I understand that members of the French Surete, who have among the Vietnamese a reputation for greater effectiveness than American intelligence officers, served three-year tours in Vietnam. -29- C. All components of US intelligence in Vietnam inspect their policies concerning in-country transfers with the purpose of decreasing their frequency. D. The OIA reassess its policies concerning the learning of the Vietnamese language by its case officers. Although I am aware that several Agency officers destined for field assignment were set to learning Vietnamese in mid-1968 (in reversal of earlier language policies) I question whether their numbers -- which I do not know -- are sufficient. 122. Short-term Recommendation Five: Steps should be taken to ensure greater cooperation between military and civilian research components in Saigon and Washington. The steps might include measures to ensure that military and civilian personnel on the analyst level can freely exchange information and opinions. The purpose of the measures would be to prevent the withholding of evidence or methodologies on which major studies are based, a practice which has happened frequently in the past. 123. Short-term Recommendation Six: An Inter-Agency Committee should be formed to review various intelligence research tools and products. Among the programs and situations which need rethinking are: A. The Hamlet Evaluation system, sound in concept, but so long misused that its statistics, as usually presented, are extremely misleading.* B. The enemy's "manpower balance" (i. e., his manpower levels, inputs and outputs). US intelligence has done such an inadequate job in earlier years concerning the enemy's numerical strength and his reserves -- both North and South -- that we now lack a firm grasp on his present capabilities. Long-term Recommendations Transcending Vietnam 124. The underlying premise of my long-term recommendations is that the overall performance of the intelligence community during the Vietnam conflict has been weak. Although some intelligence officials were uneasy *Any re-evaluation of HES ought to be accompanied by our attempt to estimate from documents the number of people under VC control, according to VC statistics, which are probably more realistic than ours. in 1964 over the possibility of a large US commitment to Vietnam, their malaise was not translated into documented exposition. They relied on "gut feelings", as did people who were more optimistic about our prospects in Vietnam. After our intervention, formal intelligence discussions of many key subjects continued to be heavily laden with unresearched supposition, and clashes between schools of thought sometimes resembled the partially informed and rambling disputes of drunks at a bar. 125. My concern over the conduct of intelligence has therefore arisen from its often slipshod nature. as has been suggested, US intelligence was inadequate in 1964 because its machinery was failing to function in certain important areas. The memorandum has demonstrated that basic questions concerning enemy manpower were hardly considered until after our intervention was a year and a half old. Unilateral espionage operations did not begin in earnest until mid-1996. Research on the enemy's backbone, the Party apparatus, has started to come into its own only recently. 126. Individual rather than mechanical shortcomings were responsible for some of there failures. In certain cases, individuals failed to turn the machinery on. In others, they neglected to retool the machine to fit the problem. Individual failure, however, is not the subject of this memorandum. 127. My long-term recommendations are largely about mechanics, and are concerned with such matters as organization and personnel policy. They are oriented primarily towards research, an area with which I am relatively familiar. They involve the Deputy Directorate of Plans only in passing. 128. Long-term Recommendation One: The Central Intelligence Agency should restructure its Deputy Directorate of Intelligence (DDI) so that it can devote more of its resources to in-depth research, particularly on political subjects. 129. The principal reason for my recommendation has been the demonstrated inadequacy of the DDI organization during the Vietnam war. One or the reasons the DDI had no one working on enemy manpower until the second half of 1967 was that no office existed to look into such matters. Likewise, the reason its reporting on the Party apparatus has been deficient is that no group of people have been designated to cover the subject systematically at headquarters.* *Technically, the Research and Analysis Staff (RAS) in Saigon, which works on Party Affairs, is a DDI component. Unfortunately, the RAS product tends to be submerged in the deluge of other reports gushing from Saigon. -31- 130. One must recognize that the DDI's main purveyor of political memoranda, the Office of Current Intelligence (OCI), seldom has time to produce in-depth research, and perforce focuses (sic) its energies on selecting and rewriting field cables concerning immediate crises.* Because of the press of deadlines, its reporting of the political activities of the Viet Cong infrastructure has usually been unsystematic and some times ill-informed. 131. In the hopes of improving the machinery, I advance the following alternatives as tentative suggestions for reorganization: A. The creation of a major new DDI component to handle in-depth political reporting, leaving "current" reporting to OCI. B. Or, alternately, a restructuring of OCI so that far fewer of its analysts are assigned to writing day-to-day material for the Current Intelligence Bulletin, or publications like the daily Situation in Vietnam. Such recurring, newspaper-style reporting could be left to a relatively small group of people (like those who write the President's Daily Brief), while analysts freed from these time-consuming chores could conduct in-depth studies. C. Or, alternately, a complete reorganization of the DDI along geographical lines, with the mixing together of the Directorate's three main substantive components: OCI, the Office of Economic Research (OER), and the Office of Strategic Research. Such a reorganization has been proposed before, and rejected, largely for administrative reasons. Although some of these reasons may be valid, I cannot help but be struck by what seems to me the duplication of effort between the two principle DDI divisions working on Vietnam.** Were they combined, the number of analysts freed might be adequate to staff a component to conduct in-depth research on the Viet Cong infrastructure. Furthermore, overall research on the war could be considerable rationalized. 132. Obviously, a battery of arguments can be marshaled to bombard nay of the alternate suggestions. To those who would resist a change in the present system, however, I would emphatically reiterate that it doesn't work. 133. Long-term Recommendation Two: The DDI should greatly increase the professionalism of its researchers. 134. The reason for the suggestion is that DDI researchers often *There are, of course, noteworthy exceptions to the generalization. **The Indo-China divisions of OER and OCI. -32-   135. To increase the professionalism and background of DDI researchers, I would suggest: A. A dramatic increase in the number of DDI personnel serving overseas. For administrative reasons, it would probably be necessary to assign them temporarily to the DDP (with the DCI footing the bill( as reports officers, or as background researchers on operational problems. (See Paragraph 139A) As a corollary and as a money saving device, I would recommend the total abolition of DDI "orientation trips" which are expensive and largely unproductive. B. An increasing emphasis on language training for DDI researchers. For example, no DDI research analyst speaks or reads Vietnamese. Because of the abstruse translation problems which have arisen concerning Viet Cong terminology, this lack has often been sorely felt. C. The upgrading of analysts within the DDI. Basically, this would involve paying them higher salaries. It would have a two-fold purpose: first, to attract better people, and second, to retain the better researchers in analytical posts. Too often the best analysts either quit or are promoted to largely administrative positions, which often means the loss of their hard-won experience. Some of the extra money spent could probably be saved by removing part of the DDI's large administrative/supervisory structure, much of which appears superfluous. 136. This recommendation presumes a basic change in attitude towards research and towards analytical personnel. In one sense, it favors the specialist over the "generalist," in that it demands of the analyst a far more rigorous performance than is usually asked for under present organizational arrangements. 137. Long-term Recommendation Three: The CIA should take steps to increase cooperation between the DDP and the DDI, bearing in mind the need for maintaining their organizational integrity. *See, for example, Robert Shaplen's "Letter from Saigon," New Yorker, II January 1969, a more perceptive discussion of recent events in Vietnam than is often found in American intelligence publications. -33- 139. Among the steps I would suggest for implementing the recommendations are: A. The creation of additional research groups abroad, similar to the DDI-manned Research and Analysis Staff in Saigon, which has proved to be remarkably useful. Obviously, most CIA stations are too small to warrant separate research components, but some of the larger ones would almost certainly benefit from them.* Were more Research and Analysis Staffs created, they should maintain close contact with country analysts in DDI headquarters. Although their day-to-day research should be for the support of the local station, they should also have the ability to service requirements from Washington. B. The temporary transfer of some DDP personnel to the DDI for two-year tours, in order to acquaint them with research problems and needs. C. The ability of the DDP to levy requirements on the DDI at headquarters for certain types of basic research. D. The setting up of procedures at CIA Stations abroad to ensure that DDI requirements sent electrically are serviced more thoroughly, and with more dispatch. A frequent -- and often valid -- complaint voiced by DDI analysts at headquarters is that cables dealing with requirements are neglected or answered inadequately. 140. Long-term Recommendation Four: The intelligence community should create on inter-Agency staff to review the history of the Vietnam war in order to develop intelligence contingency plans to avoid or to cope with future struggles of National Liberation (when deemed a threat to US interests.) 141. The principle reason for the recommendation is to help ensure that intelligence community learns and preserves the lessons that Vietnam conflict seems to be teaching us. It is advanced in the expectation that prospective revolutionaries in other parts of the world may come to look on the Viet Cong structure as an operational and organizational model. *For example, Thailand. -34- PSM page 42 of 46   Abbreviation Key ARVN Army of the Republic of Vietnam CI Counter Intelligence COMINT Communications Intelligence COSVN Central Office of South Vietnam DDI Deputy Directorate of Intelligence DD Deputy Directorate of Plans GVX Government of Vietnam HES Hamlet Evaluation System MACV Military Assistance Command Vietnam MPS Ministry of Public Security NIE National Intelligence Estimate CCI Office of Current Intelligence OER Office of Economic Research OB Order of Battle VAS Research and Analysis Staff (once called the Collation Branch)