A Fresh Look at Collection Requirements

18 SEPT 95


An unsolved problem analyzed in depth and some approaches recommended to solution.


Clyde R. Heffter

In the immediate post-war period, the word "requirement" was seldom heard in intelligence circles, and what we now know as collection requirements were managed in a very offhand way. Today this subject is well to the fore, its importance acknowledged by everyone. Looking back, it is possible to see certain steps by which this reversal of things came about.

First there was a time when many people, both collectors and consumers, saw no need for requirements at all-when information was believed to be there for the plucking, and the field intelligence officer was considered to need no help in deciding what to pluck. This period overlapped and merged quickly into a second one in which requirements were recognized as desirable but were not thought to present any special problem. Perhaps the man in the field did, after all, need some guidance; if so, the expert in Washington had only to jot down a list of questions and all would be well.

A third phase began when it was recognized that requirements were an integral and necessary part of the intelligence process and that they needed to be fostered and systematized. Committees were set up, priorities authorized, channels established, forms devised, control numbers assigned. Thus by the early 1950's the formal requirements machinery of today was mostly in place.

The fourth and most interesting phase, which is still with us, might be called the phase of specialized methodologies. The harsh difficulties of intelligence collection against the Sino-Soviet Bloc have driven home the realization that the way a requirement is conceived and drawn, the way it gets at its ultimate objective, the details it includes, the alternatives it provides, the discretion it permits, and a dozen other features may largely predetermine its chances of fulfillment.

Specialists in many fields, intent on solving immediate, concrete problems, have created new types of requirements peculiarly adapted to their own aims and circumstances. On requirement may take its shape from an analytical technique, XXX XXXX XXXX XXX. Another may be cast in the mold of a collection method-photography, ELINT, exploitation of legal travel. Subjects, areas, sources, access, communications-all have put their mark on the writing of requirements.

If we turn from the past and speculate on the future, we can hardly doubt that it will be one of intensified effort. For it is more and more evident that the answers we get are intimately conditioned by the questions we ask, and that asking the right questions--the business of requirements--is no spare-time job. But what direction should this intensified effort take?

Undoubtedly the healthy specialization and experimentalism of the present should and will continue. But by itself this is not an adequate program. The problems of requirements are not all special problems. Some of them are central to the very nature of the requirements process. One cannot help feeling that too little of the best thinking of the community has gone into these central problems--into the development, in a word, of an adequate theory of requirements.

It would be untrue to imply that nobody has been concerning himself with the broad questions. Much expert thought has gone into the revisions of guidance papers for the community at large or for major segments of it. But there is often a conspicuous hiatus between these high-level documents and the requirements produced on the working level. Dealing with general matters has itself become a specialty. We lack a vigorous exchange of views between generalists and specialists, requirements officers and administrators, members of all agencies, analysts in all intelligence fields, practitioners of all collection methods, which might lead at least to a clarification of ideas and at best to a solution of some common problems.

It is the aim of this paper to incite, if possible, such an exchange of views. It offers as candidate for the title of Number One Requirements Problem the problem of priorities. More exactly, it is the problem of how to formulate needs and priorities in such a way as to facilitate the satisfaction of needs in a degree roughly proportionate to their priorities, through the most effective use of the collection means available.

This problem is one which deserves and will probably reward the most searching study that can be given it. The present paper cannot claim to be such a study. Among its limitations is the fact that the writer's personal experience is confined to the clandestine collection field. It seeks, however, to demonstrate that there is such a general problem; that it is amenable to general analysis; that it must be examined not merely as a problem in administration but as one in analytical method; and finally that it is one with which the individual intelligence officer can effectively concern himself. The few specific proposals in the following pages are incidental to these general aims.

We may begin with a provisional definition of a collection requirement as simply "a statement of information to be collected." Our next step is to examine the most important varieties of such statements.


Kinds of Requirement

In the management of collection requirements there are certain persistent tendencies that reflect the divergent interests of the participants. There is the tendency of the analyst to publish a list of all his needs in the hope that somebody will satisfy them. There is the tendency of the theorist and the administrator to want a closely knit system whereby all requirements can be fed into a single machine, integrated, ranged by priorities, and allocated as directives to all parts of the collection apparatus. And there is the tendency of the collector to demand specific, well-defined requests for information, keyed to his special capabilities.

These tendencies are capable of complementing each other usefully if brought into reasonable balance, but their coexistence has more often been marked with friction.

It will be useful at this point to take a look at the word "requirement" in ordinary English usage. For the divergent tendencies just mentioned have a remarkable parallel in certain divergent but thoroughly ingrained connotations of the word itself. It is highly likely that these connotations, jumbled together loosely in the backs of our minds, help to create our notions of what a requirement "really ought to be." Though not mutually exclusive, they are sufficiently different that as one or the other predominates, the character of the resultant concept varies appreciably.

The first connotation is that of need. A requirement is something needed, or a statement of that need. This meaning does not necessarily involve the idea of authority. The need is objective; it is determined by the facts of the case. Thus food, water, and oxygen are requirements of the human organism. And thus information on various subjects is a requirement of the analyst, the intelligence organization, and the Government itself. When we think of the intelligence requirements of the Government, we are thinking not merely of what has been authoritatively determined to be needed, but of what actually is needed. This way of regarding requirements, which is basic and which we all share to some extent, adds dignity to our conception of our work. To the analyst, who thinks in terms of what he needs in order to do his job, it is the dominant connotation, and in fact the only one he cares about until experience forces him to look farther.

The second connotation for most people is that of compulsion or command, stemming from authority. As children we are "required" to go to school. In college we must take certain "required" courses. In intelligence, many of us regard a requirement as essentially a directive from a higher echelon to a lower one. In this view, the key question is not whether the information is objectively needed but whether its procurement has been directed by competent authority. It is a view which commends itself to the administrator, who would, of course, contend that certification by competent authority provides the best assurance that a valid need exists. This connotation, like the first, exists in varying degrees for everyone. Where it dominates, it leads to an emphasis on machinery, systems, channels, committees.

Finally, there is the connotation of request. Though "request" is no longer an active meaning of "require," both come from the same root, along with "inquire," "question," and "query." In intelligence this meaning has again come into its own. Under this interpretation, one equal (the "customer") makes a request or puts a question to another (the collector), who fulfills or answers it as best he can. There is a sort of honor system on both sides-with a dash of mutual suspicion. The requester vouches for the validity of the requirement, though the collector is free to reject it. If he accepts it, the collector gives an implied assurance that he will do his best on it, and this the requester is free to doubt. In any event the relationship is a mutual one, and in its pure form is free from compulsion. The use of direct requests appeals particularly to the collector, who finds that it provides him with more viable, collectible requirements than any other method. It sometimes appeals also to the requester-analyst, who if he finds a receptive collector is able by this means to get more requirements accepted than would be possible otherwise. Again, it is sometimes disillusioning to both, if the collector comes to feel overburdened or the analyst to feel neglected.

These three connotations of need, compulsion, and request are embodied in three kinds of collection requirement, to which we shall arbitrarily give names--the inventory of needs, addressed to the community at large and to nobody in particular; the directive, addressed by a higher to a lower echelon; and the request, addressed by a customer to a collector.


The Requirement as Inventory of Needs

An example of the inventory of needs is the series of Periodic Requirements (recently relabeled Reporting) Lists issued by the CIA Office of Current Intelligence. No collector is directed ("required") to collect against these lists; the lists are not addressed to any single collector. Some responsible individuals in clandestine collection (branch chiefs and station chiefs) have refused to handle the PRL's on the grounds that they are "not really requirements," i.e., they are not requests to the clandestine collector for information which only he can provide. In most cases, however, the PRL's are selectively utilized for guidance despite their character as inventories. There are several reasons for this. Revised three times a year, they are the most up-to-date of requirements. Their main subject, current affairs of chiefly political significance, is one which engages the interest and competence of nearly all collectors and which presents opportunities to neatly all. Many such opportunities are sudden and gratuitous; they divert no effort from other requirements, hence raise no issue of priorities.

Generally speaking, however, the inventory of needs does not appeal to the busy collector. When he accepts it, it is a sign that adequate requirements addressed to his particular capabilities are lacking. But the collector's viewpoint is not the only pertinent one. The inventory of needs can have great value as an instrument of analysis within the intelligence production office that originates it. The one thing it can not do is to contribute significantly to the resolution of the priorities problem.


The Requirement as Directive

The most broadly controlling document in the field of requirements is the list of Priority National Intelligence Objectives issued annually as a Director of Central Intelligence Directive to which attention is given in the NSC itself. Technically not requirements, and certainly not collection requirements, the PNIO's establish general guidelines for both collection and research. They are ranged in three priorities and contained in four pages. They are comprehensive, authoritative, and community-wide in their application. But because of their extreme generality, the PNIO's provide no practical guidance in settling issues of specific collection priorities. They form a constitution which requires both laws and courts to interpret it. To only a limited extent do present collection directives provide such "laws" or the USIB committee structure such "courts." It is still common practice for individual customer requirements (chiefly of the "request" variety) to claim a priority derived directly from the master document. If conscientiously applied, this practice is sound as a discipline to the requester. But it has no more value in judging the relative urgency of two specific collection requests than citation of the U.S. Constitution would have in settling a suburban zoning dispute.

On the level of collection requirements proper, the directive occurs in several situations. The clearest example is where there is a command channel, as between a collection organization's headquarters and its field representatives. Any requirement sent through such a channel is a directive if the higher echelon chooses to make it one. (Paradoxically, by euphemism, the fact that command authority is being exercised will often be indicated by the word "request.")

For purposes of this discussion, the most significant type of directive is that which emanates directly or indirectly from the authority of the DCI, or is issued in consequence of agreements between two or more agencies. Typically, such requirements originate outside the collection organization--often through the mechanism of an inter-agency committee--and represent the coordinated interests of major customers. Where requirements of this kind are traditionally and without question accepted by the collection organization and issued with command force to its components, it is reasonable to classify them as directives without looking into the precise authority of the committee concerned.

Directives are most practicable in the following circumstances: (a) where a command relationship exists; (b) where there is only one customer, or where one customer is incomparably more important than the others; (c) where a single method of collection is involved, and where this method has very precise, limited, and knowable capabilities. The last of these circumstances is most likely to occur in collection by technical methods. In such collection, especially on the Sino-Soviet Bloc, directives have been relatively successful. For when it is perfectly clear, as it often is in technical subjects, that it is possible to have this or that but not both, it becomes both feasible and necessary to reach a binding decision. In these circumstances, priorities have real meaning.

The situation is very different in some other fields where the need for priorities and hence for directives is felt equally keenly. One such field is the broad area of clandestine collection. Clandestine collection, though distinguished by its methodology, is not a single method but a congeries of diverse methods. Its capabilities are limited, but for the most part are neither precise nor knowable. The demands on it are fantastic. It serves as many customers as there are members of the intelligence community, but is under the command of no customer office. In short, it combines a maximum need for direction with a minimum of the characteristics that make direction practicable. In these circumstances the Interagency Clandestine Collection Priorities Committee which is charged with determining priority requirements for collection by the Clandestine Services of CIA, has an unenviable mission. The lists of requirements and targets (IPC Lists) issued by this body of USIB representatives since 1951 form a fascinating record of attacks on the requirements problem, from the highly selective, 18-target USSR list of 1952 to the encyclopedic, 379-target list of 1956, the subsequent selective excerpts from that list, and the worldwide list now in preparation.

The IPC Lists have served various important purposes: they have established goals, provided a basis for planning, and recorded in small compass many of the most critical information needs of the USIB agencies. The IPC has also addressed itself continuously to the problem of priorities. Its primary method has been to relate its requirements for clandestine collection to the objectives set forth in the PNIO's, and to as sign each requirement the priority carried by the corresponding objective. This method, and the variations on it, will be discussed at a later point in this paper. It cannot be said to have helped much in solving the concrete problem of deciding what items, among all items that are probably collectible, are most worth collecting at the expense of something else.


The Requirement as Request

Examples of the requirement as request can be found everywhere. Most requirements fall in this category, including a large majority of those bearing RD numbers in the community-wide numbering system administered by the CIA Office of Central Reference. The fact that RD stands for Requirement Directive is historically interesting but not currently significant.

A request may range from a twenty-word question to fifty-page questionnaire. It may ask for a single fact or thousand related facts. Its essence is not in its form or content but in the relationship between requester and collector.

An important variant on the request is the solicited requirement. Here the request is itself requested, by the collector. The collector, possessing a capability on an existing general requirement (of any of the types discussed), informs the appropriate customer of the capability and asks for specific requirements "tailored" to it. The resulting requirement is drawn up with an eye to the nature of the particular sources to be used, rather than merely to the presumed over-all capacities of the collecting organization. Through this interaction of consumer and collector, requirements of great precision and immediate practical value are developed.

In clandestine collection the solicited requirement is regularly used for legal travelers, for defectors and returnees, and for other sources whose capability or knowledgeability can be exploited only through detailed guidance or questioning. It is the cornerstone of the requirements system managed by the Interagency Defector Committee.

The solicited requirement blends into the jointly developed requirement. Here collector and consumer work out the requirement jointly, usually on a subject of broad scope and usually on the initiative of the collector. This too is a practical device of often considerable merit.

The possible variations on the request are innumerable. The unsolicited or "spontaneous" request is the basic requirements tool of the community, the means by which all can seek help from those they think able to help them. The solicited request is a precision tool for relating needs and capabilities. If capabilities were ample enough to fulfill all needs, no other form of collection requirement would be necessary. But needs are infinite, capabilities limited, priorities therefore essential, and some form of directive indispensable.


The Study of Priority

If this description of the kinds of requirement is valid, it is evident that each of the three kinds answers a deep-felt need, has a life of its own, and plays a role of its own in the total complex of intelligence guidance. Since the focus of this paper is on the problem of priorities, it must concern itself chiefly with the directive. But while the directive is the only practical vehicle for priorities, requests are also very much in the picture since priorities must govern their fulfillment.

In approaching the priorities question, it is natural to think first in terms of administration and system. Adequate administrative arrangements are in fact essential, and will be discussed in some fullness. In themselves, however, they are powerless to do more than make the wheels go round. If the wheels are also to mesh, the question must be studied further as a problem in intellectual discipline, involving analytical method and an appropriate language. Finally, it must be viewed in relation to the training and responsibilities of the individual intelligence officer. Each of these approaches will be examined in turn.


System and Administration

There exists no single, general requirements system. What might be called the requirements situation has previously been well described in this journal,1 but a brief recapitulation will be useful here.

A department or agency which engages in collection primarily to satisfy its own requirements generally maintains an independent requirements system for internal use, with its own terminology, categories, and priorities, and with a single requirements office to direct its collection elements on behalf of its consumer elements. This pattern is characteristic of the military departments. The same requirements office that performs these internal functions (or perhaps a separate branch of it) represents both the collector and the consumer elements in dealing with other agencies.

Where, as in CIA, the consumer components are dependent on many collectors and the collection components are in the service of consumers throughout the community, no such one-to-one system is possible. Each major component (collector or consumer) has its own requirements office. There may also be requirements officers at division and branch levels, as in the Clandestine Services.

Requirements offices differ in many respects, but in all cases they are the official channels for the movement of requirements between agencies. Their personnel are middlemen, and must have some understanding of the problems not only of those whom they represent but of those whom they deal with on the outside. The consumer requirements officer must find the best collection bargain he can for his analyst client; the collector requirements officer must find the best possible use for the resources he represents, while protecting them from unreasonable demands; each must restrain his own side from ill-advised intransigence.

Between agencies (or between major components of CIA) the typical requirement moves officially from analyst to consumer requirements office to the CIA Office of Central Reference to collector requirements office to collector. (Even this is a simplified statement.) OCR's community-wide system whereby such requirements are numbered and recorded makes for convenient reference. In some cases OCR also performs other functions normally performed by requirements offices, such as checking to make sure that readily available sources have been canvassed before levying a requirement on an expensive collection system.

Although the vast majority of requirements move officially through the channel just described, many of these movements are merely in confirmation of advance copies which have previously passed directly between the two requirements offices concerned. Matters of substance are regularly discussed by one requirements officer with another. And beyond this there are many instances where one or both of the requirements offices are unaware that a requirement has been agreed upon between analyst and individual collector until a confirmation copy comes through channels.

From the standpoint of the "free market," of bringing analyst and collector together, this way of doing things works well. Where the collection situation is such that effort on a low-priority target does not actually detract from the effort that can be made on a high-priority target, little harm can be done. Or where analyst and collector are both highly knowledgeable and responsible, the results can be excellent. The former condition still prevails in some areas outside the iron curtain; the latter has been attained in certain components. But neither analyst nor collector nor yet requirements officer is competent to set priorities.

Hitherto we have spoken of requests and directives as clear cut categories. But it is necessary to take account of a special variety, the request-cum-persuasion, and its still more vigorous relative, the request-cum-pressure. The intense efforts which are often made informally to induce individual in the collection offices to give special emphasis to particular requirements are a clear sign that there is a felt need for priorities. But priorities are slippery. Let us see how a typical collection priority is handled on the working level.

The OCR form used for RD's has a place for the requester to check "degree of need" as "standard," "great," or "urgent." If the analyst checks this in a way that is grossly out of line his own requirements office will probably catch him up; if i does not, the collector's requirements office will balk. But although it may be assumed that the requesting requirement office would not approve an "urgent" rating unless the requirement deserved it in relation to other requirements placed by that office on the same collection organization, no such assumption can be made as to its priority relative to requirements from other consumers. And it would be a very self-confident collector who would try to settle the question unaided.

If the collector should show no interest in a requirement marked "urgent," the requester may try proof, persuasion, or pressure. He may indeed, in anticipation of resistance, have originally indicated a relationship between his requirement and one of the Priority National Intelligence Objectives. He is almost certainly right that a relationship exists, but there may be question of its cogency. It is possible to tie a very small requirement to a very big objective. Early warning is important, but not everything described as early warning is equally important. The collector may still be unimpressed. There is no impartial arbiter, short of the USIB itself, for the requester to appeal to.

Oddly enough, in requests addressed to the Clandestine Services it is unusual for a requester to cite an IPC List. Yet in theory there should be many such citations. The Lists are designed to contain all the highest priority requirements for clandestine collection. They carry priorities derived authoritatively from the PNIO's. Moreover, taken together they are more than a hundred times longer than the PNIO's and are crammed with specifics. It would be much easier to prove that a request is significantly related to an IPC item than that it is significantly related to a PNIO-provided it actually is.

The chief reason for the paucity of citations seems to be that only a small proportion of requirements received as requests are actually on subjects specifically covered in the IPC Lists. The Lists are not cited for the simple reason that they contain nothing suitable to cite. On non-Bloc areas this is not surprising, since the IPC Lists have scarcely begun to touch them. But on Bloc areas it is astonishing, all the more since the Lists are composed of requirements and targets originally submitted by the very analysts who now make these requests for information. Is it possible that the preparation of IPC Lists is regarded by some analysts as a formal, academic exercise unrelated to the real expression of their keenest interests? Or do the Lists contain only items of such rarity and difficulty that on ordinary workdays nobody really hopes to get them? Or is it that the day-to-day requirements deal with matters so current that the IPC Lists have not caught up with them? Or with matters too unimportant to merit inclusion in the Lists?

Be the answer what it may, the fact is that the analyst in our hypothetical situation would probably appeal to a different source of authority in his effort to show the collector the importance of his requirement. The chances are good that, if he had a case capable of being pressed at all, he would draw support from positions taken by one of the substantive USIB committees that concern themselves with requirements. Among these committees are the Economic Intelligence Committee, the Scientific Intelligence Committee, the Joint Atomic Energy Intelligence Committee, and the Guided Missile and Astronautics Intelligence Committee. Each of them is authorized, among its other duties, "to recommend . . . intelligence objectives within the over-all national intelligence objectives, establish relative priorities on substantive needs, review the scope and effectiveness of collection and production efforts to meet these objectives, and make the necessary substantive recommendations to the departments and agencies concerned." Each is also authorized "to determine the deficiencies" in its own category of intelligence, "to take appropriate remedial actions, and to recommend to the Intelligence Board remedial actions" beyond its own cognizance. Such recommendations have from time to time been made and approved, with the result that priorities on very specific matters have been established by the USIB. Priorities so established have, of course, the force of directives. And such priorities have frequently been cited effectively in the levying of requirements, although the exact applicability of the priority in the context of a given collector's responsibilities has not always been clear beyond doubt.

It is the intent of this paper to illustrate rather than to exhaust the questions it raises. The requirements situation has many other significant systems and phenomena: the special, closed requirements systems governing technical methods of collection; the Watch Committee with its General Indicator List; and, to name but one more, the Critical Collection Problems Committee, whose recommendations on specified critical collection questions carry great weight with the Intelligence Board. But enough has been said to give a sense of the administrative question.

It would seem to involve at least the following aspects: (a) a vast number of requests with no consistently effective way of relating them to established directives and hence to priorities; (b) directives all of which emanate directly or indirectly from the USIB, but through different channels and without sufficient coordination, so that their impact is often disconcerting.

One sometimes encounters the view that all requirements should be fed into a single mechanism, where the marginal ones would be eliminated and the others properly related, subordinated, formulated, and allocated. The appeal of this dream is that such a procedure would, at least theoretically, deal head-on with the problems of priorities and capabilities, and would ensure that all relevant considerations and interests were taken into account simultaneously by a single responsible authority.

There are a dozen reasons why such a scheme is impractical. The group charged with this function would have to be delegated an unprecedented amount of the authority of the USIB. It would be unwieldy in size. Its staff would have to consist mainly of substantive experts and experts on capabilities, yet neither of these could be extensively spared from regular analytical or collection duties, and if away from such duties long would lose their expertise. Such a group could not keep abreast of current developments, and much of its output would be stillborn. It is inconceivable that it should take over direction of the self-contained collection systems, or that it could do so successfully. If charged with processing all requirements, its machinery would whir meaninglessly over the many that present no problem and find it difficult to pause for those that do. The priority system would probably be too standardized to help with the really hard and painful decisions. There would be a strong tendency to ratify the obvious and sidestep the prickly.

To encounter these faults and dangers, to be sure, it is not necessary to create this gigantic requirements mechanism. We face most of them already; they are the hazards of all centralized systems, whether large-scale or small-scale. But since we have still so much to learn about how to make priorities work, it would seem sensible to do our experimentation on a scale where the strains are tolerable.

One such experiment might confine itself to requirements for clandestine collection by CIA. These might be divided into directives and requests. The directives would be issued-possibly by a strengthened IPC-under new procedures to be established by the USIB. The issuing body would take advantage of the specialized competence of the substantive intelligence committees, and the latter would coordinate with the issuing body any recommendations to the USIB affecting clandestine collection priorities. All directives and priorities presented to the Clandestine Services (except those received directly from the DCI) would reach them through a single channel and would constitute a single, interrelated body of guidance. Frequency of revision would be essential. Special emergency priorities established through command channels would be possible as they are today. As for requirements of the request type, they would be served in much the present manner except that on challenge they would have to be justifies by a demonstrable relationship with a directive.


Discipline: Method and Language

The more one reflects on it, the more one sees that the setting of priorities is a singularly subtle and elusive task. Useful priorities simply cannot be caught in the coarse nets of authority, information, channels, and the division of labor. These things are needed, but so also are a disciplined intellectual approach to the subject, comprising a delicate analytical method and, perhaps most important, an adequate language.

To illustrate problems in method, we may draw once more on the experience of the IPC. That committee, as was noted, earlier, derives its priority system from the PNIO's. The system consists of three priorities, based on the degree to which the United States could be benefited by the achievement of an objective or harmed by the failure to achieve it. All IPC requirements and targets (the latter being institutions or installations on which information is needed) bear the same priority as the PNIO to which they are related. Since the PNIO's on the Sino-Soviet Bloc are all of First or Second Priority, the IPC requirements on those areas are too. The result is that a list of 300 Bloc targets may have 100 of First Priority and 200 of Second Priority.

There are several difficulties here. One, which the IPC has for some time recognized and tried to overcome, is that two priorities simply do not provide enough span. By various devices-arranging certain related targets in an internal order of importance; describing certain targets as substitutes for others; treating targets as subordinate to "basic requirements" which are sometimes expanded into several paragraphs-the IPC manages to convey a somewhat more discriminating sense of priority.

A second difficulty is that a requirement related to a First Priority objective is really not necessarily more important in itself than another requirement related only to a Second Priority objective. Everything depends on how significantly each requirement is related to its objective--how far its fulfillment would go towards achieving the objective. It is illogical to suppose that every item of information (or every target) has an importance strictly proportionate to the importance of the objective on which it bears, however minutely. Here again the IPC has recognized the difficulty and has tried to compensate for it to the extent compatible with its system. Where a requirement or target bears on both a First and a Second Priority objective, it is ranged under the objective to which it would contribute more significantly. This still leaves a tremendous unevenness in the importance of targets assigned the same priority.

Still a third difficulty is that a requirement meriting a given priority in the context of total U.S. security interests does not necessarily merit the same priority in the context of a particular collection method. The economic stability of a certain friendly country may be of great importance (Second Priority in the PNIO's), yet may not require clandestine collection at all. This difficulty also has been recognized, and where it is agreed that a requirement can be satisfied by other methods it is omitted from the List.

Unquestionably the difficulties of the priority-allocating process could be illustrated equally well from the experience of other bodies, though perhaps none faces so baffling a task. And the difficulties cited are only a few among many. These are the kinds of matters which appear much simpler before studying them than afterwards. The fact that they are nowhere near solution is one reason for keeping our experiments in priority administration on a medium scale, rather than magnify the problem by creating more grandiose structures.

In order to clarify and refine our method we need a better language. Here the most pressing need is for a common vocabulary in which such indispensable words as objective, requirement, target, and request can be relied on to mean at least approximately the same thing to everybody. This happy state can not be attained by promulgating official glossaries, but only through continued, careful discussion of common problems by persons from all parts of the community.

As we probe the more subtle aspects of requirements theory, we may find that language itself is putting blinders on us in our search for method. For instance, in the parlance of intelligence direction specific requirements are said to be "derived" from general ones which in turn are "derived" from the PNIO's or a similar authority. Is it possible that this concept of "derivation" is really no more than a convenient but misleading fiction; that the specifics are actually thought up independently and, at best, are then matched with the generalities? The same process is often described as "translating" requirements or as "breaking them down." It is not suggested that we discard such expressions but that we analyze their implications and limitations. Nobody literally believes that a PNIO of fifty words somehow contains within itself' the hundreds of thousands of specific questions that will be' asked somewhere, sometime, in the effort to fulfill it. We, know that many of those specific questions are not inevitable. Others could be substituted for them, perhaps advantageously. There is indeed a relationship between the fifty-word PNIO and the innumerable small questions, one which admittedly can never be fully charted; but has it been adequately explored? In looking into this particular matter and here we are momentarily returning from the question of language to the question of method-it would be useful to consider the history of the recently suspended specialized annexes to the PNIO's as well as of a stillborn experiment several years ago by the Office of Current Intelligence in the articulation of a body of intelligence requirements at a middle level of abstraction between the PNIO's and collection requirements.

The final aspect of the language question, and perhaps the most important, is the skill with which requirements themselves are expressed. What is needed here is not different words from those now used, but surer ways of communicating the essence of a matter from one mind (or set of minds) to another. There is no formula for this but a trained alertness to the perils of misunderstanding.


Training and Personal Responsibility

In the last analysis every action is performed by an individual; and in intelligence it is clear that the individual cannot expect to be helped more than half way by systems and methods. This is true in the field of requirements as elsewhere. To adapt a hoary but still valid epigram, requirements are far too important to be left to the requirements officers. In types of collection requiring individual initiative and judgment, these qualities must be applied to ends no less than to means. It is pertinent, therefore, to add a word about the role of the intelligence officer through whom requirements are finally put to work.

In the training of new case officers-the second lieutenants of clandestine collection-substantial attention must continue to be given to the interpreting, tailoring, questioning, soliciting, and developing of requirements suited to their sources, as well as to the training, briefing, debriefing, directing and redirecting of sources in response to requirements. The case officer must learn to study carefully the requirement which comes from far-away Washington, to grasp its purpose as well as its letter, to flesh it out with all the knowledge he has or can get, to cable for clarification when necessary, to adapt it to the understanding and the access of his sources. He must also learn to study the reporting as it comes in from I the source, and from it to develop his own immediate feedback of further questions without waiting for the customer's reaction.

To illustrate the case officer's strategic position at the crossroads of outflowing direction and inflowing product, the usual image of the intelligence cycle might be twisted into a figure 8, the upper part representing all the paraphernalia of higher echelons, the lower the collection situation for which the case j officer is responsible. He himself appears, not on the outer periphery of a vast, impersonal, revolving wheel, but where he feels himself to be-at the center, receiving and giving direction downward, receiving and submitting reports upward, himself deriving and feeding back direction from the reports he receives.

The symbolic crossroads of the figure 8 is equally applicable to the analyst in a consumer office. He too is at the center; he too must communicate upwards and downwards; he too is no cog in a machine, but a mind at work. When the systems and doctrines have been perfected, the job will still have to be done by these two.


1 By William P. Bundy in "The Guiding of Intelligence Collection," in the Winter 1959 issue (III 1), and, in the narrower context of clandestine collection, by Lowell M. Dunleigh in "Spy at Your Service. Sir." in the Spring 1959 issue (III 2).

Top of page


Historical Document
Posted: May 08, 2007 07:30 AM
Last Updated: Aug 04, 2011 02:25 PM