Leo Strauss and the World of Intelligence
(By Which We Do Not Mean Nous)
Gary J. Schmitt and Abram N. Shulsky
The topic must appear at first as a very strange one: what possible connection could there be between the tumultuous world of spies and snooping paraphernalia, on the one hand, and the quiet life of scholarship and immersion in ancient texts, on the other? However, intelligence isn't only involved with espionage and whiz-bang gadgetry; a large part of it deals with the patient piecing together of bits of information to yield the outlines of the larger picture. When one considers that this effort, called "analysis," often focuses on such major questions as the nature and characteristic modes of action of a foreign regime, then perhaps the juxtaposition of political philosophy and intelligence may seem less far-fetched. Indeed, in his gentleness, his ability to concentrate on detail, his consequent success in looking below the surface and reading between the lines, and his seeming unworldliness, Leo Strauss may even be said to resemble, however faintly, the George Smiley of John LeCarr?'s novels.
The trends in political science that Strauss polemicized against in his "Epilog" to Essays on the Scientific Study of Politics' also affected the world of intelligence. In a famous book, which laid out an agenda for the development of U.S. intelligence analysis in the post-World War II era, Sherman Kent, Yale history professor and former member of the World War TI-era Office of Strategic Services (the OSS, forerunner of the CIA) argued that intelligence analysis should adopt the social science method which was then being elaborated in the academy: Research is the only process which we of the liberal tradition are willing to admit is capable of giving us the truth, or a closer approximation to truth, than we now enjoy . . . . we insist, and have insisted for generations, that truth is to be approached, if not attained through research guided by a systematic method. In the social sciences which very largely constitute the subject matter of strategic intelligence, there is such a method. It is much like the method of physical sciences. It is not the same method but it is a method none the less.2 This method was meant to be a means of predicting the future, specifically, predicting the future course of action of a foreign government. It was applicable to any government; in a uncharacteristic bit of whimsy, Kent describes the application of his method to forecasting the actions of "Great Frusina," an amalgam of the names of the other permanent members of the U.N. Security Council, as if to emphasize that it didn't matter whether one was dealing with a constitutional monarchy, a chaotic republic, a mature totalitarian tyranny or a revolutionary dictatorship.
Kent's faith in the power of this method was so strong that he disparaged the more traditional types of intelligence information, i.e., the types of secret or "inside" information that could only be gathered by spies able to penetrate the foreign government's inner circle and/or steal its documents, or by interception techniques and codebreakers able to listen in on its communications and decipher them. As he correctly pointed out, a Soviet spy who had full run of American secret documents in the first half of 1950 could not have found one that laid out U.S. plans to defend South Korea from invasion by the North, for the simple reason that the decision to do so hadn't yet been taken, and, in fact, wouldn't be until the invasion had already begun. But this example also highlights the extremely high standard that Kent was attempting to set for the "social scientific" method: in principle, it was to be able to predict decisions that hadn't yet been made and about which the very participants in the policy process were uncertain; with this method, one could understand the decision-making process better than the decision-makers themselves.
This ambition depended crucially on the idea that, however disparate political systems may appear, the underlying political processes were universal (rooted in human nature, as it were, although a proper 1950s social scientist would have been the last person in the world to use the term). As a result, they could be discerned by an empirical method that observed behavior, tallied it, calculated correlations between particular actions and particular features of the context in which they occurred, and so forth.
Another nontraditional feature of Kent's program was that it explicitly downplayed the importance of the possibility of deception. An atomic physicist needn't be concerned with the possibility that the particles he studies are attempting to mislead him into thinking that they behave otherwise than they in fact do; and, generally speaking, social scientists can have the same confidence in their data (although, it has been suggested that, in the 1996 Israeli election, some voters vented their anger at the media by deliberately misrepresenting how they had voted when questioned by those conducting exit polls.) Given that he thought that intelligence analysis should deal with fundamental issues (such as a nation's capabilities and interests) rather than ephemera (what one of its leaders said yesterday), Kent believed that intelligence analysts could be equally unconcerned with the possibility of deception on the part of the governments they were studying. (How, after all, could Truman have deceived Stalin about his intentions in Korea if, prior to the invasion, he didn't know them himself?) While Strauss never, of course, addressed the question of intelligence analysis, it is easy to guess what he might have said about Kent's proposed methodology, since it was based squarely on the developments in social science that Strauss attacked. The primary point of attack would have been that it ignored the differences among "regimes" (or types of government and society) in its search for universal truths of social science. While Strauss was interested in understanding human nature, he understood from his study of the tradition of political philosophy-from Aristotle, most of all-that, in political life, universal human nature is encountered not in its unvarnished state, but as reflected through the prism of the "regime."
Because of the importance of the regime, it would be foolish to expect to be able to deduce theories of political behavior that would be universal, i.e., that would apply to democracies and tyrannies alike. With Tocqueville, Strauss would have argued that the regime shapes human political action in so fundamental a way that the very souls appear different. For this reason, among others, social science could never hope to be "scientific" in the sense of the natural sciences, which can be confident that the phenomena it studies do not vary from place to place. The other issue raised by Kent's methodology-the general disregard of deception-is also tied to the tendency of modern social science to submerge clear differences between various forms of rule in favor of explanations that rest on the sub-political. Although it should be obvious that some regimes are more inclined to be "open" than others, Kent's reliance on the universal aspirations of modern social science seem to have blinded him to that fact. Combined with American intelligence's great confidence in its ability to collect intelligence by technical means (space-based photographic reconnaissance satellites, ground-based listening posts, etc.), Kent's willingness to downplay the issue of deception meant that American intelligence analysts were generally reluctant throughout the Cold War to believe that they could be deceived about any critical question by the Soviet Union or other Communist states. History has shown this view to have been extremely naive.
Strauss is of course famous for his doctrine (or, rather, his discovery) of "esoteric" writing, i.e., the idea that, at least before the Enlightenment, most serious writers wrote so as to hide at least some of their thought from some of their readers. Strauss was attacked for this doctrine on various grounds. Many critics argued that it gave license for fanciful and arbitrary interpretation of texts; once one asserted that an author's true views might be the opposite of those that appear on the surface of his writings, it might seem that the sky was the limit in terms of how far from the author's apparent views one could wander. However, the deeper reason for the unpopularity of this doctrine was different; after all, Strauss was a piker compared to the very popular (at least for a while) doctrine of deconstructionism which gave readers complete carte blanche when it came to interpreting texts, and which completely lacked the rigor Strauss brought to the problem of textual interpretation. Rather, the dissatisfaction was political in origin; the notion of esoteric writing is clearly at odds with the main political tenet of the Enlightenment, i.e., that a good polity can be built on the basis of doctrines that not only are true but are also accessible: their truth can be "self-evident" (to quote the Declaration of Independence) to the average citizen. Even those post-moderns who no longer believe that it is possible to discover any truths at all on which a free polity might be based somehow still cling to freedom of speech, which was originally defended on the grounds that the propagation of anti-republican heresies can do no harm as long as prorepublican truths are left free to refute them.
Be this as it may, Strauss's view certainly alerts one to the possibility that political life may be closely linked to deception. Indeed, it suggests that deception is the norm in political life, and the hope, to say nothing of the expectation, of establishing a politics that can dispense with it is the exception.
On both of these counts, then, studying political philosophy with Strauss proved to be a valuable counterweight to the doctrines that were then prevalent, not only in the academy, but in intelligence analysis as well. By emphasizing the distinction among regimes as the basic political fact, political philosophy prepared one for a much better understanding of the world than did the "scientific" social science which sought to understand the various regimes in terms of universal categories. As many observers have noted, a characteristic failing of American intelligence analysis is what is called "mirror imaging," i.e., imagining that the country one is studying is fundamentally similar to one's own and hence can be understood in the same terms. As described by Eliot Cohen,
A far more serious problem ... centers on the possibility that if policymakers read [estimative] intelligence, it will mislead them or reinforce inappropriate prejudices. The official school of intelligence writing seems to pay very little heed to problems of deception and concealment, a serious deficiency in view of the premium placed by many regimes ... on such activities. But more pervasive, and even more pernicious, is the phenomenon of mirror imaging by intelligence analysts .... It is a varied and subtle phenomenon and can afflict those who pride themselves on their hardheaded realpolitik as much as it does those who take a sunnier view of international relations.'
This fault shows up in many ways. Cohen cites, for example, the use of the terms "moderates" and "extremists" to describe the various participants in Iranian political life in the 1980s. While there may well have been an internal struggle going on in Iran, use of these terms was misleading. For example, the term "moderate" would imply someone who wanted better relations with the West and who favored a relaxation of the rules enforcing strict religious practices; however, there is no reason why, in the Iranian context, someone holding the former view should also be expected to hold the latter. (Clearly, as Americans used the term, "moderate" meant nothing more than "more like us": but this is obviously a ridiculous category to use when trying to understand a very different society.) As Cohen points out, "That bloody 'extremist' Robespierre initially opposed a warlike foreign policy, as did the no less radical Lenin."4
Mirror imaging also affects the judgments of intelligence analysts concerning how foreign officials think about the strategic problems they face. Cohen cites a number of cases when assuming that foreign leaders who think about these matters in the same way as Americans proved disastrous. The problem takes an almost comic turn in the following defense of a 1962 intelligence estimate that incorrectly assessed that the Soviets would not put missiles in Cuba:
In that case, as Sherman Kent often said, his estimate of what was reasonable for the Soviet Union to do was a lot better than Khrushchev's, and therefore he was correct in analyzing the situation as it should have been seen by the Soviets.5
Many reasons are cited why this, particular problem should be so deeply rooted in American intelligence analysis: the failure of our educational system to teach foreign languages; a "universalistic" outlook which believes (not entirely incorrectly) that others aspire to an American way of life; the "melting pot" tradition, which suggests that, despite superficial differences of language, customs, etc., people are fundamentally alike and want the same things. While these are all plausible contributors, the influence of American social science may be an even more important and deeper cause. The study of political philosophy and its emphasis on the key importance of the variety of regimes is an important antidote.
Similarly, many critics of American intelligence have noted that it tends to ignore "open sources," in particular, what foreign leaders say about their beliefs and intentions. While one must be alert to the possibility of deception, one must nevertheless start with the "surface," as Strauss would have put it. The careful reading of what foreign leaders say would be an obvious beginning point for understanding what they really think, even though the two should never be considered as simply identical. For example, at the time of the Iranian revolution in 1979, it appeared that the intelligence analysts at the CIA did not have easy access to Khomeini's writings about religion. In part, this reflected the standard social science view that, in a modernizing society such as Iran, religion was destined to play an increasingly minor role. (A reading of Thucydides' account of the role that religious passion played in causing the failure of Athens' Sicilian expedition would have sufficed to guard against that particular mistake: Athens was clearly the most "enlightened" of ancient Greek cities.) But it also reflected the view that one could assess the views of a Khomeini from the outside, without having to try to understand him as he understood himself. Strauss's painstaking method of recovering the thought of thinkers of previous times would have been applicable to understanding someone like Khomeini, whose intellectual world was so different from our own.
With the end of the Cold War, the struggle of ideologies has come to a close. Some have foreseen an "end of history," in the Hegelian sense of the attainment of philosophic self-awareness; others, a "clash of civilizations," in the sense of the conflict of what are ultimately mutually incomprehensible value systems. For those brought up in the realist tradition, it will seem strange that theories of international relations should have such philosophic origins and implications. Nevertheless, such is the world we face; and the study of the classics of political philosophy with Leo Strauss was a surprisingly good preparation for grappling with it.
NOTES 1. Herbert J. Storing, ed. (New York: Holt, Rinehart and Winston, 1962). 2. Strategic Intelligence for Americans World Policy (Princeton: Princeton University Press, 1949), , p. 155. Kent's influence on the goals and methods of intelligence analysis in the United States was not limited to the fact that his book became the best selling book of its kind and a bible of sorts for those within American intelligence. After the U.S. was surprised by the invasion of South Korea in 1950, Kent was asked to join the Office of National Estimates (ONE), a new intelligence unit charged with producing comprehensive, forward-looking intelligence assessments, thereby hopefully precluding further surprises of this sort. Soon afterwards, Kent became the director of ONE-at the time, the senior analytic post in U.S. intelligence; he held that position for more than 15 years. As Bruce Berkowitz and Allan Goodman note, "ONE and the process of developing NIEs [National Intelligence Estimates] bore a strung resemblance to the principles for analysis Kent described in Strategic Intelligence?' (Strategic Intelligence for American National Security, [Princeton: Princeton University Press, 1989], , p: 5)
Kent's views on the application of the scientific method to intelligence analysis also influenced the institutional arrangements for carrying it out. Even before World War II had ended, Kent and other analysts within the OSS's analytic arm had reached the conclusion that the positivist approach to analysis-resting on Max Weber's fact-value distinction-should be reflected institutionally in a sharp division between intelligence analysis and policy-making organizations. Historically, foreign intelligence analysis in the U.S. and elsewhere had been located in government departments directly responsible for carrying out the key national security functions of war or diplomacy. Under the new ethos of social science objectivity, however, scholarly distance was essential for intelligence analysts. Thus, the CIA is not part of a policy-making department of government and is located geographically in the Virginia suburbs, away from the White House, the Pentagon, and the State Department.
Reflecting the general disillusionment among social scientists themselves with respect to the predictive capabilities of modern social science, scholars studying past intelligence "failures" began to question the assumptions and utility of Kent-like theories of analysis in the late-1970s. However, only in recent years has the intelligence community itself begun to challenge Kent's views and their hold on the practice of intelligence analysis. See, in particular, Douglas J. MacEachin's "The Tradecraft of Analysis" and Joseph S. Nye's "Estimating the Future" in US. Intelligence at the Crossroads, Roy Godson, Ernst R. May and Gary Schmitt, eds. (McLean, VA: Brassey's, 1995), pp. 63-96.
3. "Analysis," Intelligence Requirements for the 1990s, Roy Godson, ed. (Lexington, MA: D. C. Heath and Co., 1989), pp. 76-77.
4. Ibid., , p. 77.
5. Discussant remarks of Ray S. Cline in Roy Godson, , ed., Intelligence Requirements for the 1980s: Analysis and Estimates ('Washington, D.C.: National Strategy Information Center, 1980), p. 77. Cited in Cohen, op. cit., , p. 78, emphasis supplied.