Cloak & Gown: SCHOLARS IN THE SECRET WAR by Robin W. Winks (Morrow: $22.95; 596 pp., illustrated)

Robin W. Winks, master of Berkeley College at Yale, has really written two different books under the same cover. The first one is a completely original study of how a great university, Yale, marshaled its intellectual resources during World War II to help the government organize its intelligence services. The second one is an intellectual history of some of the men who helped shape American espionage--notably, James Jesus Angleton, Donald Downes and Norman Holmes Pearson. Surprisingly, the two books combine neatly to provide one of the best studies of intelligence in recent years.

The strength of this book is that Winks approaches espionage as a historian rather than as an investigative journalist. The difference is a difference in the disciplines. Instead of relying primarily on interviews, he goes, wherever possible, to written source material--personal papers, letters and even (in Angleton’s case) grade sheets at Yale--and then, through interviews, clarifies its meaning. Moreover, when dealing with secondary sources, he critically assesses them, pointing out to his readers their strengths and weaknesses. Finally, and most critically, instead of concealing his sources, as too many journalists do (a practice that allows them to pass off copied work as original), he devotes a 120-page section of his book to spelling out the sources of his information and the circumstances under which he obtained it. The result is a book that not only gives meticulous order to the chaos of anecdotal accounts about espionage but also puts it all within the universe of verifiable knowledge.

The issue that interests Winks is the historic role of academic men of knowledge in the dark discipline of intelligence-gathering. Since Yale has been a main recruiting ground for American spies--as well as the repository for the collected papers of many who served in this capacity--he begins by investigating the role its alumni played in America’s first centralized intelligence service, the OSS (which stood both for Oh-So-Social and Office of Strategic Services). Gradually, as he interviews more than 200 persons involved in such intelligence work, his study broadens to cover the intellectual development of U.S. intelligence through 1975.

Unfortunately, by focusing his attention on some of the more fascinating characters in American intelligence and their social and academic milieu, Winks blurs over some important conceptual distinctions. For example, he confuses HUMINT--which is the abbreviation in the CIA for data dependent on a human spy--with human intelligence itself, which he then defends by asserting that “an effective intelligence community require(s) effective human beings.”

Winks would oppose man to machine, machine being represented by SIGINT (which is the abbreviation for electronic signals intelligence). What he misses is that this abbreviations refers only to the source from which this data, intercepted from foreign nations, comes--not to how it is evaluated. SIGINT comes from intercepted signals, HUMINT from foreign nationals acting as spies, RADINT from enemy radar transmissions, PHOTINT from aerial cameras over foreign territory. The capitalized letters notwithstanding, all these modes use humans--and human intelligence--to research, analyze and evaluate the captured data.

The issue that divided the CIA was thus not that of humans versus machines but that of a dangerous, old-fashioned, illicit mode versus a safe, modern, licit mode; that is, of espionage, euphemistically called HUMINT, versus “national technical means,” including satellites and remote antennae “vacuum-sweeping” electronic transmissions. Espionage, by definition, is illicit because it depends on inducing foreign nationals, in violation of the laws of their country, to photocopy or steal information. National technical means are legal because they are sanctioned by the SALT treaty and attractive because the satellites or antennae that intercept the data are not in enemy territory.

To an intelligence bureaucracy the difference between the two is considerable. Espionage is more likely than national technical means to provide arrests, defections or embarrassing incidents. On the other hand, an enemy can anticipate national technical means--can anticipate, that is, the path of a satellite or range of an antenna--but can never be certain what data its bribed or disaffected officials may be giving away through espionage. And precisely because espionage is, when successful, unanticipated, Angleton and others in the CIA maintained that, despite its dangers and illegality, espionage (HUMINT) was a crucial means of insuring that the mass of information photographed or otherwise sucked in by the all-too-predictable and all-too-well-understood technical means was not being used to deceive us.

Unlike a scientist peering through his microscope, who need not worry that nature is out to dupe him, a counterintelligence officer has to consider the possibility that what he “sees” has been intentionally placed in his path to mislead him. In the final analysis, the CIA debate came down and still comes down to an epistemological question: How can we know that the information we gather has not been made available to dupe us?

While Winks has not answered this question--and neither, to my knowledge, has the CIA--he has provided the best intellectual background on the men at the center of the debate. And, on this ground alone, I would recommend “Cloak & Gown” to anyone seriously interested in the thinking that goes into the secret war.