Sherman Kent was perhaps the foremost practitioner of the craft of analysis in American intelligence history.(1) One of the architects of the US Intelligence Community, for nearly two decades he was at the head of its estimative machinery. He wrote one of the first books to systematically address the topic of intelligence analysis, a work that is still assigned reading in college courses and still stands as perhaps the most lucid of its kind. (2) Throughout his career he strove to imbue the intelligence discipline with a sense of professionalism, and as part of that effort he helped to found the CIA's professional journal (3) and inspired the formation of its Center for the Study of Intelligence (CSI).
On the other hand, as extensive as were his experience and analytical acumen, Sherman Kent's career was not unblemished. Just before the Cuban Missile Crisis he and his colleagues in the Office of National Estimates committed one of the greatest estimative blunders in the history of the US Intelligence Community. In 1962, the Intelligence Community not only failed to foresee the Soviet deployment of offensive missiles to Cuba, but also in September also published a Special National Intelligence Estimate (SNIE) that argued against any such prospective deployment. As head of the Board of National Estimates, Kent was responsible for this judgment and an active participant in the discussions that produced it. Soviet offensive missiles were discovered in Cuba only because the Director of Central Intelligence (DCI), John A. McCone, chose to ignore the advice of his intelligence analysts and pressed for U-2 reconnaissance flights over the island. Typically, Kent later held up his own error as a lesson to be learned by future generations of intelligence analysts and emerged from the experience with his reputation and integrity enhanced. When he retired it was rightly seen as the end of an era.(4)
Born in Chicago in 1903, Sherman Kent spent much of his childhood in California and Nevada. In 1911 his father, William Kent, was elected to the first of three terms as a California congressman and from then until 1919 Sherman lived in Washington, DC. He attended Yale University, graduating in 1926 and receiving a Ph.D. in history in 1933. In 1941 Sherman Kent was rescued from what (one suspects) ultimately would have proved to be an unsatisfying teaching career at Yale by the imminent entry of the United States into World War II.(5) Along with other rising intellectual talents, Kent joined Research and Analysis Branch (R&A) of the newly created Coordinator of Information (COI), later to become the Office of Strategic Services (OSS). In January 1943 he became Chief of R&A's Europe-Africa Division. After the war, when the OSS was dissolved, Kent did a brief stint in the State Department's Office of Research and Intelligence (built on the bones of R&A) before returning to academic life. In late 1950 he joined the Central Intelligence Agency (CIA) to become the deputy chief of the newly created Office of National Estimates (ONE), working for his old boss at the OSS Research and Analysis Branch, the Harvard diplomatic historian William L. Langer.(6)
Together, Langer and Kent were to preside over a reorientation of the analytical components of the CIA. That this was needed was due less to any intrinsic flaw in the Agency's conception than to its inability to cope with the bureaucratic environment in which it existed. The CIA had been created as a national intelligence organization with capabilities that transcended and complemented those of the established intelligence units in the State Department and the military. The national intelligence produced by the CIA was intended to be a synthesis of the vast, often highly technical, knowledge possessed by the agencies that made up the US Intelligence Community,(7) which would be disseminated in a form easily accessible by the policymaker.(8) By its very existence, the CIA thus challenged the authority of the existing bureaucratic structure. No sooner was the Office of Reports and Estimates (ORE, the early CIA's analytical arm) established than it found itself balked on all sides. Internally, the intelligence process was hampered by bitter disputes over the proper coordination of analysis,(9) while externally ORE only managed to alienate the defense establishment as a w hole by reporting on the limits to Soviet military power (thereby contradicting much of the rationale behind the funding of a large peacetime military) and by persistently bursting the bubbles of anti-Soviet hysteria that appeared during this volatile period.(10)
Forces hostile to the new intelligence organization as it then existed began to mobilize within a year of its foundation. In 1948 the newly formed National Security Council (NSC) established the Dulles-Jackson-Correa Committee to investigate the performance of the Agency, with results that were devastating to ORE. On 7 July 1949, the NSC accepted the committee's report and recommended the massive reorganization and reform of the CIA. The DCI, RAdm. Roscoe Hillenkoetter, resisted implementing the report's conclusions, but the final blow was struck less than a year later, after ORE failed to foresee the outbreak of the Korean war on 25 June 1950. After North Korea invaded the South, President Truman nominated Lt. Gen. Walter Bedell Smith, US Army, to replace Hillenkoetter as DCI, with a mandate for a complete overhaul of the agency. (11) "Beetle" Smith, who had been Eisenhower's Chief of Staff during the war, had known Langer as head of the Research and Analysis Branch of OSS. In November 1950, Smith brought both Langer and Kent into the CIA to put its analytical house in order.
This was more than a bureaucratic reshuffling. It was at once a reaffirmation of the principles that underlay the creation of a national intelligence organization and a rejection of the experiences of the CIA's first three years. The old Office of Reports and Estimates was broken up and replaced by the Office of National Estimates, with a research arm in the Office of Research and Reports. The intelligence reports produced by ORE were replaced by a new form of finished intelligence, the National Intelligence Estimate (NIE).(12) Responsibility for producing the NIEs lay with the Board of National Estimates, part of the Office of National Estimates. At the interagency Intelligence Community level, a supervisory Intelligence Advisory Committee reviewed the quality of the analysis, ensured that it represented the opinions of all participating intelligence organizations, and guaranteed the right of dissent for those that disagreed with the conclusions of an NIE, all or in part. Langer became the head of the Office of National Estimates and chairman of the Board of National Estimates, with Sherman Kent his deputy in each. When Langer returned to Harvard in early 19 52, Kent succeeded him in both positions, where he remained until he retired at the end of 1967.
Under this leadership, the Board of National Estimates emerged as a body with the authority to issue even unwelcome substantive intelligence judgments, while remaining tolerant of dissent from within the ranks of the "departmental" intelligence organizations.(13) Frankly elitist in its conception, the Board recruited many of its members from ivy-league university faculties, with a leavening of members from the military and the ranks of the professional intelligence officers.(14) Clearly, Sherman Kent's model for the estimative intelligence process was strongly shaped by his prewar academic experience, and under his tutelage the Board of National Estimates seemed to have taken on the atmosphere of a faculty common room.(15) The notoriety the Board thus earned as a haven for "liberal" east coast intellectuals may have reinforced its prestige and reputation for intellectual integrity in some circles, but in the end this reputation left it isolated in the intensely political environment of the Washington defense and foreign policy establishment. This level of detachment (or, if you will, isolation) was enhanced when the CIA moved across the Potomac to its new Langley, Virginia, headquarters compound in late 1961. More or less secure in its authoritative position as the sole source of estimative intelligence in the 1950s, the Office of National Estimates was increasingly under fire throughout the following decade. Finally, faced with the overt hostility of the Nixon administration, it was unable to defend itself effectively and in 1973 was reorganized out of existence.
Nevertheless, for the best part of a generation the Board of National Estimates stood at the pinnacle of the American intelligence establishment, for the most part with Sherman Kent at its head. His influence upon the growth of intelligence analysis, both as a process and as an institution, was at once far reaching and pervasive. It is difficult to imagine a man more qualified for this job than Sherman Kent. This California-bred, Yale-trained historian turned intelligence officer combined superb academic training with an instinctive grasp of the problems at hand. Moreover, his no-nonsense, shirt-sleeve style could convey often highly speculative syntheses in a form congenial to the pragmatic minds of policymakers and military men alike. Kent's concern for the intelligence process, however, went beyond the immediate task of producing intelligence reports to a consideration of the nature of intelligence itself, to a passionate interest in the growth of intelligence analysis as a profession, and to its establishment as a scholarly discipline with a well-ordered methodology.
In thinking about intelligence, Sherman Kent nonetheless began with an understanding of national power that was well within the mainstream of contemporary American strategic thought. Kent's contribution was to apply thinking about strategy and national power to an ordered conception of intelligence analysis as an intellectual discipline. At its most fundamental level, his work began with implicit recognition of strategic intelligence (perhaps better thought of as strategic intelligence analysis) as a social science. Strategy, by definition, is a plan to achieve some given end. If we think of a national security strategy (or policy) as a blueprint for preserving the life and health of a nation, then there must be some idea of what that nation is: why does it function, how does it function, and what is essential to its survival? Conversely, a strategy to confront another nation in conflict--be it in war or cold war--must consider the strengths and weaknesses of a potential opponent in detail. Central to both lines of thought is a general conception of national existence: what comprises the nation-state, what makes it strong, and what is necessary to its survival and prosperity.
Strategic intelligence, to use Kent's term, thus takes as its subject the sinews of national life. In its most elevated form, it considers the nation-state to the depth and breadth of its being. This is what Sherman Kent called a nation's strategic stature: not just the means it possessed to wage war, but its total potential for war--the resources that are available, or might be made available; the population, industrial plant, and transportation net; the political and social structure, their stability, and the "moral quality of the people and their strength of values"--their willingness to be mobilized for war and the reasons for which they would fight--and, lastly, the political leadership, their strength and "genius (or want of it) for organizing men and materials into a community of life and strength."(16)
War, of course, is the most extreme form of confrontation between nation-states and (particularly in the nuclear age) one that is rationally considered only as a last resort. As the most fundamental component of its strategic stature, however, a nation's war potential could be seen as a benchmark for measuring its response to other, less violent, forms of confrontation and interaction. In the immediate postwar world, with the global strategic balance polarizing into opposing power blocs, such considerations took on a whole new importance. Moreover, one global war having just ended, the possibility of another such confrontation did not seem remote in the middle of this century. Indeed, the prospects for global war permeated American strategic culture and consistently stood uppermost in the minds of policymakers.
Thus, even though Kent's conception of strategic intelligence was built "in a context--and hope--of peace," its principal focus remained "the possibility of war and the mobilization of armed power."(17) In peacetime, he wrote, strategic intelligence did not differ "in any fundamental sense" from its wartime equivalent, but could be distinguished only by differences in "emphasis and direction." Whereas the strategic analyst in peacetime would concern himself with as-yet unmobilized war potential and general questions of policy and strategy, in wartime the "big question" became that of an enemy's full-developed military capabilities and their potential use in pursuit of specific strategic objectives. In an age of total war, "nearly all the factors of war potential" were still very much present in the wartime strategic calculations, but actively applied in a situation in which the life and health of the nation are measured by its ability to maintain "armed force . . . at maximum operational activity, without undue damage to overall strategic commitments, without overstraining or ruining the home war economy, and without shattering the staying power of the polity and society."(18)
In essence, Kent viewed strategic intelligence with the eye of the historian: focused on long-term trends, he tended to think of crises as "blips" in broader historical continuities. With Kent at the head of the CIA's estimative machinery, a system was created that was better adapted to studies in military economics or Soviet strategic forces planning than to plotting sudden shifts in the Soviet political leadership or to problems of strategic warning. Perhaps more important, the system established under Kent largely equated national intelligence with strategic intelligence. The result was an analytical world view that was at once comprehensive (since it viewed war potential as the ultimate expression of national power in all its forms) and curiously two-dimensional, concentrated as it was largely on the military aspects of the problems of national security. American intelligence analysts became encyclopedic in their knowledge of politics and economics, in the peoples and countries of the globe, but war, or rather the potential for war, remained their bedrock concern, especially with reference to the Soviet Bloc. From the Berlin blockade to the fall of the Soviet regime, the potential for all-out military confrontation with the USSR ran like a red thread through the fabric of US strategic intelligence, sometimes fading, sometimes growing in saturation and intensity, but always coloring the American perception, not only of the Soviet Union, but of the world in general. While perhaps appropriate to Cold War analyses of the Soviet "threat," it was a perspective that left the Intelligence Community less well prepared to deal effectively with the quite different questions that dominate the post-Cold War period.
It is perhaps inconceivable--or at least unlikely--that the CIA could have played any other role in Cold War intelligence analysis, but in retrospect this role appears to have contributed significantly to the complexity of national intelligence and to have virtually guaranteed that National Intelligence Estimates were prepared in an atmosphere charged with political energy. Balanced as it was between the policymaker on the one hand and the military leadership on the other, the Office of National Estimates could maintain its credibility as an intelligence producer only by building a reputation for disinterested objectivity--a characteristic that was bound to be subject to close scrutiny by intelligence consumers. The situation was complicated, moreover, by the nature of estimative intelligence--to use Kent's terminology--which was unprecedented in history. Highly predictive in nature, the typical NIE consisted largely of informed judgments about future actions or situations that ultimately could not be proved, however well-founded they might be in experience or factual or theoretical knowledge. "Estimating," wrote Sherman Kent, "is what you do when you do not know."(19) The amount of speculation contained in an Estimate would vary widely, depending upon the subject matter, the amount of evidence available, and the scale of time involved. Inevitably, however, this kind of analysis would conta in some portion of error, either because of a mistaken judgment or because the available evidence was misleading, incomplete, or just plain wrong. In general, therefore, NIEs would try to convey the full range of possibilities, even though they might come down firmly in favor of one particular set of conclusions. Even so, at bottom, the credibility of any given Estimate depended less on the quality of the analysis that it contained than upon the general credibility of the institution that produced it.
Under these circumstances, Kent recognized that the Office of National Estimates could be effective as an intelligence-producing organization only if it purported to be nothing more than it actually was (or perhaps ought to be): a group of scholars who were prepared to consider a situation dispassionately, objectively, and thoroughly; to study a topic with the intention only of coming to an understanding of it; and then to communicate their findings dispassionately, objectively, and concisely to the people who needed to act upon them. Intelligence, he wrote, was only one of "a score of forces at work" in national policy formulation. The policymaker might choose to listen to his intelligence advisers, or he might yield to the influence of any of a number of foreign and domestic interests and pressure groups. With these, no intelligence organization could ever hope to compete; rather, the intelligence officer's goal should not be to influence the policymaker's judgment, but to be relevant within his own area of competence and to be credible:
Let things be such that if our policy-making master is to disregard our knowledge and wisdom, he will never do so because our work was inaccurate, incomplete, or patently biased. Let him disregard us only when he must pay greater heed to someone else. And let him be uncomfortable--thoroughly uncomfortable--about his decision to heed this other. (20)
Thus, in the intelligence officer as scholar, Kent saw the means of dissolving the inherent tensions in the often delicate relationships between the policymaker, the military commander, and the intelligence producer. With no equities to protect save his own credibility, the intelligence officer could afford to make his pronouncements in the confidence that his own disinterestedness would protect him from charges of undue political influence. The concept of "disinterested objectivity" lay at the heart of Kent's thinking about intelligence organization and colored ONE's relationship with successive administrations.
In an atmosphere in which interest is influence, the notion that an intelligence officer could be truly "detached" perhaps betrays a 'naivetè born of wishful thinking. At bottom, however, Kent was enough of a bureaucrat to see that ONE's reputation for dispassionate objectivity was something more than an end in itself. In Kent's view, the "national" status of ONE's finished intelligence made it a check on the aberrations of the individual intelligence services as well as the synthesized view of the Intelligence Community as a whole. One of the factors that distinguished the CIA from the departmental intelligence organizations was its ability to segregate analysis from other considerations (usually budgetary) that were external to the intelligence process. This was particularly true of the various military intelligence organizations where (he suggested) intelligence judgments were at least in part determined by the programmatic interests of the parent service. Thus, from Sherman Kent's perspective, it was difficult to dissociate Air Force judgments concerning the size of the Soviet bomber force in the 1950s from that service's own policy, which favored a large inventory of B-52s, while "for some time, it was the view in certain high quarters of the US Army that all-out conventional war between NATO and Warsaw Pact forces need not escalate to a nuclear war." "Indeed," wrote Kent,
the consequences of being the first to use nuclear weapons would be so horrendous that Army Intelligence, knowing that the US would not do it, was willing to estimate that the Soviets would likewise refrain. Hence, an estimate allowing for the contrary invited an Army objection. To draw a permissible inference is to point out the obvious. If postulated armed conflict, say, in Europe could lead to the sort of large-scale fighting of World War II and with conventional weapons, the Army had a very good reason (budgetary, doctrinal, pride of service) to keep pressure for a full strength ground force. Per contra, an estimate that held that small, conventional wars between the nuclear powers would inevitably and perhaps speedily escalate to all-out nuclear conflict (largely the mission of the USAF) would be virtually to estimate the US Army out of business.(21)
If Kent was forced to conclude from this that not all service intelligence judgments grew "out of a selfless love of country," he also acknowledged the importance of an underlying conflict of world views. In a statement that was meant to apply to the military services in general, he noted that, "In their scale of values, first came the country and second, the force necessary to protect and preserve it. They could not question the necessity of the latter, given their high-minded patriotism."(22) Since each service had its own idea of what constituted the necessary force, the intelligence process thus became an intellectual dialectic: no synthesis was possible without a thesis and an antithesis. If the task of the service intelligence organizations was to protect departmental interests and to put forward departmental viewpoints, that of the Office of National Estimates was to produce intelligence that did justice to each of those viewpoints while retaining an overarching national perspective.
To preserve this aura of Olympian detachment, Kent concerned himself with the requirement for what might be called a realistic epistemology of intelligence analysis. No doubt this arose in part from Kent's concern for the growth of intelligence analysis as a profession and his desire to establish standards and systems by which analytical knowledge could become cumulative. He wanted to create what is now referred to as an institutional memory, so that each generation of analysts would not have to relearn the lessons of its predecessors.(23) At the same time, he doubtless believed that systematizing intelligence was necessary to establish the authority of the National Intelligence Estimate. In this regard, Kent's c all for the creation of an "Institute for the Advanced Study of Intelligence," modeled after the Princeton institute, was as transparently elitist as it was far-sighted in its understanding of the enlightened interests of the profession.(24)
Like the wise men at the Institute for Advanced Study at Princeton, those in Sherman Kent's institute would earn their keep by writing about their specialties, all of which would concern the doctrine, theory, or history of intelligence. Their output would be published on its own or in the CIA's professional journal, Studies in Intelligence, and would form the core of what Kent called a cumulative literature. It was here that Kent pinned much of his hopes for the future growth of the profession of intelligence as a well-ordered discipline, and, not surprisingly, Kent himself made some of the most perceptive contributions to Studies in Intelligence (most of which have been collected in this volume). Yet in the final analysis, although he tried to make the best of the results, one suspects that they were something less than he expected. His intelligence institute was never founded, while, in its absence, the classified journal, Studies in Intelligence, proved to be too dependent on harried and frequently overworked line intelligence officers to achieve anything like the consistently scholarly quality that he had hoped for.(25)
Intelligence analysis in CIA never achieved an explicit, broadly based, epistemological and doctrinal structure. For this, Sherman Kent himself must shoulder much of the blame. Like most historians of his generation, Kent was uncomfortable with theoretical constructs, preferring in their stead empirical judgments that were founded in an ordered methodology. This is not to say that his work lacked a theoretical structure, but that it was buried in the form of an implicit set of assumptions that comprised a comprehensive world view. Thus, when Kent asked rhetorically where "such disciplines as chemistry or medicine or economics" would be "if their master practitioners had committed no more to paper than ours," he was referring to the accumulation of empirical knowledge and methodological experience from one generation to the next, rather than the synergism between empiricism and theory that in fact drives these disciplines.(26)
Kent's methodological writings in Studies in Intelligence reflect the essentially empirical framework that he applied to analysis. His essay, "Estimates and Influence," for example, combines a tidy summary of Kent's conception of the place that intelligence should occupy in the national security establishment with a masterful dissertation on the nature of estimative intelligence. Yet it provides only a superficial consideration of the epistemology of intelligence (in its discussion of the difference between "known," "knowable," and "unknowable"), even though such considerations are essential to his basic concepts of analysis. In "Words of Estimative Probability," Kent attempted to impose a charmingly useless vocabulary structure on an otherwise cogent discussion of the need for clarity of expression. Finally, although the articles in which he relates his own experiences as an intelligence analyst and producer all are informative, entertaining, and in their own way insightful, they contribute little to the fund of cumulative knowledge. The exception to this generalization is Kent's reflection on his own near-catastrophic failure in the Cuban Missile Crisis, where his self-analysis brought transcendent insight.
All this is only to say that Kent was a part of the time and place in which he lived. Kent has been criticized for failing to pay more attention to social science theory in his thinking about intelligence.(27) Yet since he came to intelligence analysis with the background and training of an historian, reinforced by his work with other historians in World War II, it is not surprising that he did not focus on the social sciences. In his contributions to the methodology of intelligence he helped to lay the pragmatic foundation upon which a doctrine and theory of intelligence analysis could be built, and he cannot be blamed if the practitioners that followed him failed to build on this foundation.(28)
In the end, the fate of the Office of National Estimates hinged not on its credibility, or authority, or even its methodology, but on its ability to focus on the interests of the policymakers it was supposed to serve. The reverse of Sherman Kent's coin of detached objectivity was irrelevance. The substantive judgments of ONE's intelligence product were always subject to criticism, but by the late 1960s it was not the quality of the research that went into NIEs that was subject to attack, nor even necessarily the validity of the judgments they contained, but rather the essential usefulness of the intelligence product itself. The assault on the Office of National Estimates reached its height when another scholar, National Security Advisor Henry Kissinger, having complained about having to "fight his way through 'Talmudic' documents to find their real meaning," worked to circumvent ONE entirely and began producing intelligence reports in the National Security Council.(29) In 1973 a new DCI, William Colby, dissolved ONE and founded the system of National Intelligence Officers that exists today.
The question that brought down the Office of National Estimates, that of how closely to tailor the output of the intelligence process to meet the interests of the policymaker (as opposed to what the intelligence producers themselves believed to be important), antedates the creation of the Intelligence Community itself.(30) Although Sherman Kent wrote extensively about the need to cater to the requirements of the policymaker--especially in Strategic Intelligence--one former analyst has noted that those "who served under Kent . . . have to take care to distinguish between the Kent of the book and the practicing Kent."(31) To the question, "How seriously should we in intelligence take the indictment which damns our estimating work as unnecessary, or misleading, or irrelevant?" Kent himself responded that while a misleading estimate was indeed damned, whether an estimate was unnecessary or irrelevant was not necessarily best determined by the recipient--thus implying that in Kent's view the Board of National Estimates was best equipped to judge these issues.(32) At some point, the Office of National Estimates crossed the line between scholarly objectivity and intellectual arrogance.
Sherman Kent's personal hour of reckoning came in 1962, when the Soviet Union decided to deploy nuclear-armed ballistic missiles to Cuba. The Board of National Estimates not only failed to predict this action (or anything like it), but argued resolutely against its likelihood, right up to the moment that offensive missiles were discovered in Cuba. On the face of it, the system created because of the strategic surprise that began the Korean war had functioned no better than its predecessor.
No one who remembers those days thinks of the Cuban Missile Crisis as other than a period in which the world was on the brink of nuclear war, when the slightest miscalculation could have pushed either side over the edge. At the time, this sense of crisis was only heightened by the American people's extreme sensitivity to any shift in the strategic balance (whether apparent or real) as well as by the Kennedy administration's preoccupation with Cuba. Under these circumstances the actual American response was one of the least extreme that might have been expected.
In retrospect, the Soviet decision to deploy missiles to Cuba now seems an act of desperation, a frontal assault on the American position in the Western Hemisphere that was unlikely to produce any lasting military advantage and, once discovered, was bound to result in the kind of direct confrontation that did ensue--one that could well have escalated to war. In fact, it was part of a high-stakes game of prestige and influence played by the Soviet premier, Nikita Khrushchev, whose motivations probably owed more to Soviet internal politics than to any reasoned strategic calculation.
It was also the kind of short-term perturbation in the strategic balance that played directly to the weaknesses of Sherman Kent's system of national intelligence. As such, it lay outside the world understood by the Board of National Estimates, in which power was built on a solid base of military-industrial strength and statesmen acted on the basis of cool calculation and measured response. Thus, when first detected in July 1962, the Soviet military buildup in Cuba was interpreted as defensive in character, the appropriate response to the previous year's disastrous Bay of Pigs invasion--an action that perhaps ought to have warned Sherman Kent and his colleagues of the folly that was possible in a Soviet-American confrontation over Cuba. As it happened, the Board of National Estimates was reinforced in its preconceptions by the fact that offensive missiles were not deployed to Cuba until very late in the game. They were preceded over a three-month period by surface-to-air missiles (SAMs), coast-defense missiles, KOMAR-class missile boats and Il-28 BEAGLE light bombers, all of which fit neatly into the concept of a defensive buildup in Cuba.
At bottom, however, it was the lack of precedent for these Soviet deployments in the measured, action-reaction sequence of Cold War confrontation that led to the Board's failure to recognize the situation for what it actually was. Thus, although in September 1962 Special National Intelligence Estimate (SNIE) 85-3-62 patronizingly conceded that "the USSR could derive considerable military advantage from the establishment of Soviet medium- and intermediate-range ballistic missiles in Cuba," such a development "would be incompatible with Soviet practice to date and with Soviet policy as we presently estimate it." Moreover, this estimate continued:
It would indicate a far greater willingness to increase the level of risk in US-Soviet relations than the USSR has displayed thus far, and consequently would have important policy implications with respect to other areas and other problems in East-West relations. (33)
In the end it was the Director of Central Intelligence, John A. McCone, who somehow tracked the aberrant course of Khrushchev's thought processes, kept the Intelligence Community on track, and perhaps staved off war. In a bizarre series of telegrams, written while he was on honeymoon in France, McCone badgered his Agency with warnings of Soviet missile deployments. Returning in September, he demanded that reconnaissance flights against Cuba be resumed, despite the presence of ever-more numerous SAMs on the island and despite the continued lack of evidence of anything beyond a defensive buildup. Finally, on 14 October, U-2 photographs revealed Soviet SS-4 medium-range ballistic missiles (MRBMs) in the early stages of deployment in sites about 50 nautical miles to the west-southwest of Havana.(34)
"There is no blinking the fact that we came down on the wrong side," Kent wrote two years later. "When the photographic evidence of 14 October was in, there was the proof." There was also no question that the Office of National Estimates had misread Soviet intentions throughout the summer and fall of 1962; yet if Sherman Kent was willing to acknowledge in retrospect that there were in fact "a handful" of indicators that ought to have warned the Board of National Estimates what the Soviets were about, he nonetheless stood by the intelligence judgments that were made. The problem, he argued, was the faulty judgment on the part of the Soviets, who had themselves misread US intentions with respect to Cuba in the months following the Bay of Pigs. In any intelligence estimate, he argued, the "other man's" decision will have been shaped by his diplomatic missions and intelligence service, which will have observed and reported the "things he must know prior to his decision."
These and other phenomena very considerably narrow the area of a foreign statesman's choice, and once thus narrowed it is susceptible to fairly sure-footed analysis by studious intelligence types. As long as all the discernible constants in the equation are operative the estimator can be fairly confident of making a sound judgment. It is when these constants do not rule that the real trouble begins. It is when the other man zigs violently out of the track of "normal" behavior that you are likely to lose him. . . . No estimating process can be expected to divine exactly when the enemy is about to make a dramatically wrong decision. We were not brought up to underestimate our enemies.(35)
At bottom, of course, Kent was correct: it is the quality of erratic and irrational behavior that it is erratic and cannot be predicted--one cannot know the unknowable. Yet one can detect the conditions and circumstances that give rise to erratic behavior and warn against it. Herein lay the intelligence failure, the same kind of error, one might argue, that allowed the United States to be surprised at Pearl Harbor in December 1941. It was to avoid this kind of mistake that a national intelligence capability was created in the first place.
"I find myself wondering a lot about Dobrynin," Kent somewhat wistfully observed, referring to the Soviet Ambassador to the United States. Was it possible that Anatoly Dobrynin, who had full access to the government and press of the United States, went along with the estimate that led the Soviets to deploy missiles to Cuba? Could he have failed to sense the "special place" that Castro's Cuba occupied in American foreign policy thinking? Could that sense of place not have been communicated to the right people in Moscow? Perhaps, Kent suggested, "we do not know some things about Soviet foreign policy decision-making that we should."(36)
During the crisis, Sherman Kent and other senior intelligence officers were sent out to the capitals of America's European allies, to explain what had happened, to explain why what had happened had not matched the intelligence that we had communicated to them, and to explain what the United States was doing to resolve the crisis, and why. Kent was sent to Paris, Brussels, and Ottawa, and, as was to be expected, he wrote up his experiences, to add them to the cumulative knowledge of the Intelligence Community.(37)
Sherman Kent remained the head of the Office of National Estimates for another five years following the Cuban Missile Crisis, retiring on 29 December 1967, after more than 30 years of government service. The system of national intelligence that he had built survived him by about five more years. Over the course of the 1960s the Intelligence Community grew and expanded its capabilities. Responsibilities shifted. In 1961 the Defense Intelligence Agency (DIA) was created, and by the end of the decade ONE, once at least nominally the hub of the Intelligence Community, was little more than first among equals. In 1973, DCI William Colby abolished the Office of National Estimates and replaced it with the system of National Intelligence Officers that persists to this day. By the 1980s ONE was held up to new analysts as an example of an organization that had lost its way, by producing intelligence that was too "academic" in orientation (whatever that means), and by failing to pay adequate attention to the needs of the policymakers who relied on its judgment. No doubt these accusations were at least in some measure true. Yet one cannot but wonder whether the Office of National Estimates had lost its way, or whether the environment in which it existed had not changed to the point where it was simply not wanted. After all, ONE was nothing more nor less than one more government institution, and institutions must change with the times. If the counsel of wise men was not easily accepted in the 1960s, it was perhaps even less welcome in the decades that followed.
By seeking to provide more relevant or "engaged" intelligence analysis, the system that replaced the Office of National Estimates brought new problems: charges of "politicization" replaced those of policy irrelevance, while the effort to keep up with current reporting increasingly interfered with the kind of in-depth research that drove the estimates. To complicate matters, the end of the Cold War has raised diverse new issues, while the fluidity of the global balance of power and the changes that seem to occur almost daily in the international arena have brought a new conception of intelligence, one that owes more to journalism than to the scholarship in which the National Intelligence Estimates were conceived. In this environment, Sherman Kent's detached approach to estimative intelligence seems stolid and archaic--much as the medieval scholars must have appeared to the new men of the Renaissance. Yet we should allow none of this to influence our judgment of the Office of National Estimates in its time, of Sherman Kent's contribution to the methodology of intelligence, or of his formative role in the growth of intelligence as a profession.
Donald P. Steury
(1) Throughout this essay, the terms intelligence and intelligence analysis will be used interchangeably. The reader will be aware, however, that the word intelligence properly encompasses the collection of important information as well as its digestion and analysis.
(2) Sherman Kent, Strategic Intelligence for American World Policy (Princeton, NJ: Princeton University Press, 1949).
(3) Studies in Intelligence, the CIA's professional quarterly, from which a number of the articles in this collection come.
(4) During the crisis, Kent helped brief the details to NATO (see "The Cuban Missile Crisis of 1962: Presenting the Photographic Evidence Abroad," in this volume) and later published a frank appraisal of what went wrong in Studies in Intelligence (his article "A Crucial Estimate Relived," is also included in this collection). An excerpt from the pivotal SNIE and other important documents may be found in Mary S. McAuliffe, ed., CIA Documents on the Cuban Missile Crisis, 1962 (Washington, DC: Central Intelligence Agency, 1992).
(5) Kent's field was electoral politics in 19th-century Bourbon France. His first book, Electoral Procedures Under Louis Philippe, was published by Yale University Press in 1937. Upon retirement, Kent resumed writing in his specialty, the results appearing as The Election of 1827 in France (Cambridge, MA: Harvard University Press, 1975).
(6) In fact, Kent did not return to teach at Yale until 1949: in 1946-1947 he taught at the new National War College and in 1947-1948 he received a Guggenheim Fellowship to write Strategic Intelligence. For a brief account of Sherman Kent's life and career, see Harold P. Ford's "A Tribute to Sherman Kent" in this volume.
(7) The "intelligence community" is made up of the intelligence gathering and reporting agencies of the US Government: the CIA, the FBI, the National Security Agency (NSA), the State Department's Bureau of Intelligence and Research (INR), the intelligence departments of the military services, the Department of Energy (originally the Atomic Energy Commission), and, later, the Defense Intelligence Agency (DIA). Although the term Intelligence Community did not come into use until the mid-1950s or later, it is used here to describe the collectivity of US intelligence from World War II onward.
(8) The CIA was founded as a "national" intelligence agency with access to all available intelligence sources and information, as opposed to a "departmental" agency (such as the Office of Naval Intelligence), which studied only those reports that were relevant to its specialized field. After World War II, there was widespread belief that the Japanese attack on Pearl Harbor could have been foreseen had there been a single intelligence authority with access to all the information that existed in diverse places before the war. For an enlightened discussion of this situation see Harold P. Ford, Estimative Intelligence: The Purposes and Problems of National Intelligence Estimating (Lanham, MD: University Press of America, 1993), pp. 3-25.
(9) For details, see Arthur B. Darling, The Central Intelligence Agency: An Instrument of Government, to 1950 (University Park, PA: Pennsylvania State University Press, 1990); and Ludwell L. Montague, General Walter Bedell Smith as Director of Central Intelligence (University Park, PA: Pennsylvania State University Press, 1992). These two volumes are declassified versions of in-house histories written from classified sources by the CIA History Staff.
(10) For example, the "war scare" of 1948. Relevant documents (among others) are ORE 1 Soviet Foreign and Military Policy, 23 July 1946; ORE 1/1 Revised Soviet Tactics in International Affairs, 6 January 1947; ORE 22-48 Possibility of Direct Soviet Action During 1948, 2 April 1948; and ORE 22-48 (Addendum) Possibility of Direct Soviet Action During 1948-49, 16 September 1948. All have been declassified and are in transit to the National Archives.
(11) For the Smith tenure as DCI, see Montague, General Walter Bedell Smith as Director of Central Intelligence.
(12) "Finished intelligence" is intelligence reporting in its final form, as it will be submitted to the policymaker. Generally speaking, finished intelligence is derived from a number of sources and is analytical or estimative in content.
(13) The intelligence organizations of the Department of State and the military, which (at least in theory) concerned themselves with the interests of their individual departments, as opposed to the overarching "national" concerns of the CIA.
(14) For example, of the 13 members of the Board of National Estimates in 1964-1965, only one had a professional military background--although six others had wartime military experience in some capacity and five more had wartime government service. One, Sherman Kent, had wartime analytical experience in the OSS. Four had Ph.D.'s; two had law degrees; and one, an M.D. Four attended Yale; four, Harvard; two, Princeton; three, Oxford; and two studied in Paris. Two were Rhodes scholars. Eight had faculty or academic experience--in most cases enough that the CIA qualified as a second career. For no one was this more true than for Sherman Kent himself.
(15) Kent's model of the "scientific method" as the basis for intelligence production perhaps came out best in his "Need for an Intelligence Literature" in this volume.
(16) HS/HC-7 CIA Progress Report; Office of National Estimates (ONE) Section I "Intelligence and the Problem of National Foreign Policy," 26 December 1951, p. 2. (This document has not yet been declassified.)
(17) Strategic Intelligence, p. 61.
(18) Strategic Intelligence, pp. 62-63.
(19) Sherman Kent, "Estimates and Influence," Studies in Intelligence (Summer 1968). See p. 35 in this volume.
(20) "Estimates and Influence." See p. 34 in this volume.
(21) Sherman Kent, The Law and Custom of the National Intelligence Estimate, Miscellaneous Studies, CIA History Staff, February 1976. See p. 82 in this volume.
(22) Ibid. See p. 82 in this volume.
(23) In the Intelligence Community, a "generation" of analysts might be as brief as four years, as analysts are trained and then pass on to other tasks, or are absorbed into management. For Kent's discussion of cumulative knowledge, see "Valediction" and "The Need for an Intelligence Literature," in this volume.
(24) Sherman Kent, "Valediction," Studies in Intelligence (Winter 1968). See p. 24 in this volume.
(25) The current Center for the Study of Intelligence, which was at least partly inspired by Kent's concept of an Institute for the Advanced Study of Intelligence, was not established until 1975.
(26) Sherman Kent, "The Need for an Intelligence Literature," Studies in Intelligence (September 1955). See p. 15 in this volume.
(27) Jack Davis, "The Kent-Kendall Debate of 1949," Studies in Intelligence (Summer 1991).
(28) As in the Davis article, supra. In fact, intelligence analysis as a profession now owes much more to the social sciences than to history. Most intelligence analysts in CIA are trained not as historians but as political scientists or economists, while those who work on technical topics are trained in the sciences and engineering.
(29) Lawrence Freedman, US Intelligence and the Soviet Strategic Threat (Princeton, NJ: Princeton University Press, 1986), p. 49. There were, of course, other matters at issue, not least ONE's position on Vietnam, which was at odds with that of the administration.
(30) Davis, "The Kent-Kendall Debate of 1949."
(31) Davis, p. 92.
(32) "Estimates and Influence." See pp. 41-42 in this volume.
(33) Mary S. McAuliffe, ed., CIA Documents on the Cuban Missile Crisis, 1962 (Washington, DC: Central Intelligence Agency, 1992), p. 93.
(34) The standard works on the Cuban Missile Crisis are Graham T. Allison, The Essence of Decision: Explaining the Cuban Missile Crisis (Boston: Little, Brown; 1971) and Robert Smith Thompson, The Missiles of October: The Declassified Story of John F. Kennedy and the Cuban Missile Crisis (New York: Simon and Schuster, 1992). No study of the crisis would now be complete without close consideration of the intelligence aspects, for which Mary S. McAuliffe, supra, is invaluable. See also her article in the Society for Historians of American Foreign Relations Newsletter (June 1993, with a reply by Samuel Halpern in the December 1993 number of the same journal); Dino A. Brugioni, Eyeball to Eyeball: The Inside Story of the Cuban Missile Crisis (New York: Random House, 1991); and James G. Blight and David A. Welch, On the Brink. Americans and Soviets Reexamine the Cuban Missile Crisis (New York: Hill and Wang, 1989).
(35) Sherman Kent, "A Crucial Estimate Relived," Studies in Intelligence (Spring 1964). See p. 185 in this volume.
(36) Ibid. See pp. 186-187 in this volume.
(37) Sherman Kent, "The Cuban Missile Crisis of 1962: Presenting the Photographic Evidence Abroad," Studies in Intelligence (Spring 1972). See pp. 189-208 in this volume.