Coruscation

Then, as now, theorists and practitioners of intelligence sought a smoothly functioning, highly efficient and seamlessly integrated organization, or cluster of organizations. But they struggled at it, largely because the purposes to which intelligence were put were complex and at times contradictory.

In his book "Cloak and Gown," published in 1987, the Yale historian Robin Winks pointed out, "The 'intelligence debate' was framed in 1949." That was the year a classic text, Sherman Kent's "Strategic Intelligence for American World Policy," came out.

To Kent, the best intelligence-gathering was the work "of devoted specialists molded into a vigorous production unit," who prized the arts of data accumulation and nonideological analysis.

Kent's book was widely adopted by intelligence services around the world. But it also had critics, among them the political scientist Willmoore Kendall, a onetime adviser to the C.I.A. He wrote that Kent's approach, influenced by the Pearl Harbor attack, betrayed "a compulsive preoccupation with prediction, with the elimination of 'surprise' from foreign affairs."

This was a worthy goal in wartime, Mr. Kendall said, but in peacetime the most useful intelligence provided the big "pictures" of the world that decision makers needed for formulating broad policy. Intelligence experts therefore should not just acquire and analyze information; they should interpret it as well.