Intelligence Analysis Demystified


Is strategic intelligence – the comprehensive analysis and assessment of international relations (in political, economic and security terms) – a matter of “craft” and “intuition”? Or is there a cognitive methodology in intelligence analysis, which approximates scientific standards like reliability and validity? 61 years ago, the American historian and intelligence analyst Sherman Kent wrote a book making the case for the latter. This “old” book deserves the status of mandatory reading for serious journalists, private sector analysts or political scientists.
by Michael Liebig

Occasionally, one has eerie experiences with books. More or less by chance, one comes across an old book which seems somehow interesting. One glances through some pages and instantly knows: This book I must read. I recently had such an experience. The book’s title is Strategic Intelligence for American World Policy by Sherman Kent, Princeton University Press, 1951 (1949).

The title of the book may raise a few eyebrows. And a Google search will reveal that Kent worked for the CIA. But Kent’s book will be of no interest for those who are into “secret intelligence” or “covert ops.” Its author is a history professor, who also looks like a history professor. Kent’s subject is the methodology of analyzing strategic affairs.

Therefore, outside the realm of state intelligence services, Kent’s book should interest people who do serious journalism – genuine “news analysis” and/or investigative journalism. Or people who work in private intelligence which does not (or should not) involve illegal, clandestine activities or “dirty tricks.” Or political scientists who do country or regional studies.

Kent’s book is an inspiration for people who do analytical work with open sources: print or electronic information in the public domain (with no legal restrictions excepts copyright) like newspapers, journals, books, radio, TV – and the Internet. In intelligence jargon, open source intelligence is “OSINT”. And, let’s not forget that state intelligence services draw at least 80% of their intelligence input from OSINT. Kent’s definition of OSINT is “unromantic open-and-above-board observation and research“ (p.4)

Who was Sherman Kent?

Kent (1903-1986) was a Yale professor of European History. In World War II, he joined the “Office of Strategic Services”(OSS). The OSS was a first attempt to “centralize” US intelligence capabilities, even though the intelligence activities of various government departments and the Army and Navy were continued parallel to the OSS. The OSS, under its chief “Wild Bill” Donnovan, is mostly portrayed as an “action”-oriented intelligence organization. But OSS developed an outstanding analytical capacity, by recruiting in particular first-class academics – historians, economists, political scientists, sociologists or geographers – to its Research and Analysis branch. In the latter, Kent served as the head of the Europe-Africa Division. Strategic Intelligence is Kent’s summary of his wartime experience at the R&A branch of OSS.

I would think that Kent had the R&A branch of OSS in mind, when writing: “In a sense, intelligence organisations must be not a little like a large university faculty. They must have the people to whom research and rigorous thought are the breath of life, and they must accordingly have tolerance for the queer bird and the eccentric with a unique talent. They must guarantee a sort of academic freedom of inquiry and must fight off those who derogate such freedom by pointing to its occasional crackpot findings.” (p.74) And Kent adds: When such freedom of thought is missing, intelligence analysts “will be alike as tiles of a bathroom floor – and about as capable of meaningful and original thought.” (ibid.)

When writing about talented eccentrics, Kent might also have thought of some leading German émigré political scientists. Among the staff of R&A were Herbert Marcuse, Franz L. Neumann and Otto Kircheimer. (For many, the “Frankfurt School” is still a controversial issue – being blamed for all sorts things, notably the 1960s “student revolt” and its consequences. My sense is that now, at last, an objective and sober judgement on the Frankfurt School is gaining the upper hand. Many of the Frankfurt School’s most fervent critics have never read any of the works of its leading representatives; doing so, might lastingly discourage prejudices.)

For Kent, “intelligence is a simple and self-evident thing. As an activity it is the pursuit of a certain kind of knowledge; as an phenomenon it is the resultant knowledge […] And strategic intelligence, we might call knowledge upon which our nation’s foreign relations, in war and peace, must rest.” (p. VIIff.)

Intelligence, writes Kent, “can be thought – indeed it often is – as an organisation engaged in the manufacture of a product (knowledge) out of raw materials (all manner of data) and labor (highly skilled, but not practical in the business sense of the word). The product, to be worthy of the label, must be up to standard… Like many a producer of consumer’s goods, intelligence will have its greatest marketing success when its products bears the unmistakable signs of superior research, cautious development, sound design and careful production.“ (p.76)

Intelligence, University & Newspaper Desk

But an intelligence service’s analytical department (other than the department operationally collecting information), is not really an academic institution. At least in the ideal case, academic research knows no time pressure; what counts is not timeliness, but scientific reliability and validity. In international relations, however, time does matter. World politics is dynamic. Multiple actors, each with diverse intentions, capabilities and time-tables, are determining the course of events. Their interaction may trigger a chain reaction, further accelerating the course of events and leading to ruptures in the strategic configuration. (I would assume that “surprise” is a dubious category for Kent; he would probably see “surprise” – at least on the strategic, other than the tactical level – as a consequence of bad analysis.)

Because time does matter so much in international relations, Kent draws parallels between an intelligence service and (quality) journalism.

“But that intelligence organizations bear resemblance to a university faculty is not enough. They must be geared to a quicker pace and must be more observant of deadlines though this may occasionally and regrettably involve a sacrifice in accuracy. Intelligence organizations must also have many of the qualities of those of our greatest metropolitan newspapers. After all, many of their duties have close resemblance to those of an outstanding daily. They watch, report, summarize, and analyze. They have foreign correspondents and home staff. Like the newspaper they have their privately developed hot sources; their speedy and sure communications. They have their responsibilities for completeness and accuracy – with commensurately greater penalties for omission and error. They have the same huge problem of handling the news in millions of words per day and seeing that the right staff man gets all the messages which fall appropriately into his field. They even have the problem of editorial control and the difficulties of reproduction and dissemination. In these terms it is fitting that intelligence organizations put more study upon newspaper organization and borrow those phases of it which they require.” (p.75)

Three Dimensions of Intelligence Analysis

The parallels between quality journalism and intelligence are the closest in what Kent calls “current-reportorial” intelligence, which nowadays is called “current intelligence”. The core quality current-reportorial intelligence is “a high capacity to detect the significant and a high sensitivity to changes.” (p. 70) This means “spotting the unusual, the really unusual,” identifying “the three things per week of the thousands it observes and the millions that happen which are really of potential moment.” (p.161). However, current-reportorial intelligence, “without its accompanying research will produce spotty and superficial information.” (p. 155)

In turn, current-reportorial intelligence is based on “basic descriptive” intelligence. The latter, writes Kent, follows the “baedeker” approach: Basic geographic, historical, demographic, economic, and socio-cultural data. That is what we find today, for example, in the – unclassified – CIA World Fact Book. Without knowing these basic data, current events cannot be adequately analyzed. And both basic descriptive and current-reportorial intelligence are the indispensable precondition for what Kent calls „speculative and evaluative“ intelligence. The latter means making an intelligence “estimate.” The German term would be Lagebeurteilung, which inherently contains prognostic inferences, for which Kent uses the English term “Estimate of the Situation.”

Kent writes about the three dimensions of intelligence: „As the reporting element carries out its task it constantly adds freshness to the content of the basic descriptive element. It does more than this, for in keeping otherwise static knowledge up-to-date it maintains a bridge between the descriptive and what I have called the speculative-evaluative elements – a bridge between past and future.” (p.38)

For Kent, intelligence estimates are the culmination of the intelligence process. Not surprisingly, after having left OSS for history teaching and research back at Yale University, Kent joined the CIA in 1950, where he was the head of the Office of National Estimates (ONE) from 1952-1967. In the mega-organization CIA, the Office of National Estimates had a staff of just 70 persons. “Are so-called ,estimates‘ of intelligence of any value?,” asks Kent, “my answer is Yes, they are of very great value if they are soundly based in reliable descriptive data, reliable reporting, and processed from careful analysis.” (p.60)

What makes an Intelligence Analyst

For Kent, intelligence is first of all a process of generating knowledge. His interest is focussed on intelligence as a cognitive process. Nowadays, the intelligence literature seems fixated on organizational and procedural formulas for the optimal “intelligence cycle.” Sixty years ago, it seems, Kent was already worried about the disregard for the cognitive dimension of intelligence in favor of such formulas: “I hold that you cannot produce knowledge of a high order of subtlety and utility in the same way you produce Fords.” (p.109)

The requirements for the intelligence analyst set by Kent are indeed demanding:

  • “The very highest competence in one or more of the sciences of politics, economics, geography, and the military art. He should not undertake it unless he has an easy familiarity with the literature and techniques of the relveant disciplines.” (p.48)
  • Intelligence analysis „requires of its producers the best professional training, the highest intellectual integrity and a very large amount of worldly wisdom.” (p.64)
  • ”Lastly, he [the intelligence analyst] must be no passive receiver of impressions. He must continually be asking himself embarrassing questions. He must be imaginative in his search for new sources of confirming or contradicting information, he must be critical of his new evidence, he must be patient and careful in ordering the facts which are unchallengeable, he must be objective and impartial in his selection of hypotheses – in short, although his job is not primarily a research job, he must have the qualities and the command of the techniques of the trained researcher.“ (p.70)

One may add that similar, maybe not quite as rigid demands would apply to intelligence-related occupations in the private sector like quality journalism. On the products of intelligence work, Kent observes: “There is such a thing as a complicated idea; there is such a thing as so complicated an idea that it cannot be expounded in 250 words, or in two pie-charts, an assemblage of little men, little engines, and three quarters of a little cotton bale.” (p.176)

Hypothesis Building & Intuition

Kent’s exploration of the cognitive dimension of intelligence analysis also takes up the crucial question of hypothesis building. When making an “estimate of the situation,” synthesizing hypotheses is an integral part of the cognitive process. But as important as generating hypotheses, is the thorough testing of hypotheses – and not the hasty and obstinate fixation on one hypothesis. The latter is often camouflaged as “intuition”. Obviously, intuition does exist and can open up an important pathway towards the articulation of hypotheses – provided these hypotheses are then thoroughly tested. Kent speaks of running “an obstacle race with his [own] and other people’s hypotheses.” (p.172)

If self-critical “testing” is discarded, he who has been begotten by intuition merely engages in what Kent calls “communion with his intuitive self.” (p. 156) His conclusion: “What I do wish to reject is intuition based upon nothing and which takes off from wish.” (p.203) And, “without discarding intuition as invariably a false friend, … use it with a full knowledge of its frailties.” (p.206) Not intuition, but “the mind is the best long-run solver of unknowns.”(p.204)

Kent provides an undeniable example for the “climatic errors” that result from discarding this warning: “Adolph Hitler had his hunches and the first few of them were brilliant… But the trouble was that he apparently did not try to analyze the why of his successful intuition. He went on as if his intuition were a natural, personal and infallible source of truth”. (p.204)

The second deadly sin of the intelligence analyst is his forgoing intellectual integrity by submitting to – Kent uses the German word – “kaempfende Wissenschaft”. In Nazi Germany, this term described the subordination of “science” and the prerogatives of the regime’s ideology. For Kent, the ultimate degeneration of the intelligence analyst is his becoming “the unabashed apologist for a given policy rather than its impartial and objective analyst.” (p.200)

To avoid any misunderstanding, I would emphasize that intelligence analysis is not “passive” and merely descriptive in respect to social reality. Even “pure” analysis is always cognitively interwoven with goal setting and purpose. That’s unproblematic as long as there is a self-reflective understanding of what “is” and what “should be.” That is the “is-ought problem” – the Sein-Sollen theorem in German classical philosophy.

There is no unbridgeable dichotomy between “idealism” and sober analysis. If you want to change the world, you first have to know the world. When “will” substitutes or ideologically deforms analysis, the resulting practice is prone to end in disaster of some kind.

Kent’s Strategic Intelligence is a groundbreaking contribution for the theoretical understanding of intelligence analysis. It’s a true “classic” of intelligence analysis. And that’s why I think that in this case extensively quoting him is indeed appropriate. Obviously, Kent’s book leaves many issues of the methodology of intelligence analysis unresolved, but it certainly stimulates further investigations into these issues. And, Kent keeps common sense: “One last word: intelligence is bound to make mistakes.” (p. 194)

I personally regret very much that I didn’t read Kent’s book at a much earlier point of time – after all, it was first published in 1949. Whatever, better late than never.

Die Kommentarfunktion für diesen Beitrag wurde beendet.