Beautiful Minds

alph Peters, a former Army intelligence officer, once located a 12-man military unit hiding in an 1,800-square-mile expanse of wilderness simply by staring at a map. He summoned his powers of logic and strategy, applied his deep knowledge of the terrain and his gut instinct, and "just knew" where the men were hiding, he says.

Peters' unconventional approach is not uncommon among intelligence analysts who have proved prescient about many of the threats the United States now faces, including global terrorism, well before the Sept. 11 attacks exposed the danger for all to see.

Yet, two years after the Sept. 11 attacks, U.S. intelligence agencies still rely on practices that thwart or water down insightful analysis, critics say. Congressional reports and inquiries into lapses in intelligence have criticized the agencies as vast, befuddled bureaucracies. Most critiques of the CIA and a dozen other intelligence agencies have said they must become more "modern" organizations, better equipped to fight terrorists. The agencies are used to outsiders judging their performance, often harshly. But when any organization truly changes, it is usually from within and not according to the prescriptions of critics.

Because that is so, intelligence analysts, especially the most creative ones, offer the best hope that American spy agencies will find ways to better understand and prevent terrorism. Fixated for the past half-century on the Soviet Union and its vast, lumbering bureaucracy, most analysts only recently have had to think like the nimble, shadowy terrorists they now face.

But some among them long have practiced unconventional thinking. They are the beautiful minds of the American intelligence community. Like Nobel Laureate mathematician John Nash, whose struggle to express his brilliance was portrayed in the 2001 film A Beautiful Mind, these intelligence analysts see the world in a way others can't. They walk a thin line between what is known and the visions they weave from bits and pieces of information gathered everywhere.

But insightful analysis hasn't been widely embraced. The intelligence community sometimes is stifled by a compulsion to speak with one voice. The customers of intelligence agencies-elected officials and senior leaders in federal agencies-want harmony, not dissonance. Often, only the least conventional thinkers can tune out the noise and find the true notes.

BLACK ART

"Great intelligence work moves into the realm of irrationality," says Peters, who served in military intelligence from his late 20s to his mid-40s. Over the course of his career, he earned a reputation as a brilliant analyst-and a maverick.

That's because Peters, to the consternation of some of his bosses, saw intelligence not as a science so much as a "black art," a cerebral tradecraft in which logic and intuition meet, sometimes with phenomenal results.

Skeptics questioned Peters his entire career. Superior officers were suspicious of how he intuited meaning from intelligence that stumped other analysts. "How do you know that?" they demanded as Peters divined Soviet war plans or predicted troop movements.

Peters rarely could fully explain himself logically. And he still can't. The best analysts are like concert musicians, he says, whose talents are innate, and seem inexplicable. But the intelligence system is threatened by these beautiful minds. "A really good analyst is bound to be the odd duck," Peters admits. And the intelligence agencies, like other bureaucracies, don't embrace deviants.

Perhaps that's because maverick analysis often appears implausible. As a historical example, Peters points to the way purely logical analysts might have viewed the bubonic plague that swept Europe in the mid-14th century.

They would have predicted that the Black Death would devastate European civilization. Indeed, it killed one-third of the population. But less conventional thinking might also have seen the plague as the opening to reshaping the entire social order of the continent. In fact, that's what happened. The plague altered social structures and ushered in the Renaissance, which gave birth to modern Europe.

Peters says intelligence agencies are intolerant of heretical analysis because it's bureaucratically threatening. But sometimes, intuition, though imprecise, proves to be right.

Peters was one among a cadre of military and civilian analysts who, in the early and mid-1990s, warned of the emerging threat of Islamist fundamentalism to the United States, overseas and domestically. No one could foretell to the day when terrorists would attack, but analysts warned of the danger and even named potential targets. Leaders of the intelligence agencies largely dismissed these predictions, which appeared mostly in speeches and government journals, as being outside the realm of possibility. Fundamentalist religious leaders, with their inferior military technology, were labeled "mad mullahs," with more bark than bite.

Just because intelligence agencies haven't embraced non-mainstream thinking doesn't mean they've produced bad work, Peters says. But they haven't produced great work. "Bureaucracies give you dependable competence, but they are the enemies of excellence," he says.

Are intelligence agencies afraid of being brilliant? Probably not. It's more likely that the managers who oversee analysts, especially maverick ones, worry about looking competent in their bosses' eyes.

Laurie Broedling, an organizational psychologist and human resources expert who worked as an associate administrator at NASA in the 1990s, says that mangers don't "bring bad news up the chain" or report "dissonant, divergent" information to their superiors if it makes them look as if they can't control their staff. But this can keep the most shrewd analysis from reaching senior decision-makers who need it most.

Can only odd ducks produce top-notch analysis? Are the best analysts born with beautiful minds?

Peters contends that superior analysts can be trained. "It's not some kid sitting over a crystal ball," he says of his own analytic style. "Talent is essential to play at the top of the game, but experience-the 'training' of direct experience of foreign environments-is indispensable." Peters says young analysts should be sent for long periods to the countries they study to expose them to the "foreign reality," learn the language and make personal connections. Immersion "engages and draws out the talent, or intuitive ability."

Arguably, the intelligence agencies have done less of that than was needed, and not in the right places. After a decade of personnel and budget cuts, the CIA struggled to find analysts who could speak the languages of the Central Asian nations that are now the focus of the war on terrorism. Some ex-CIA officers have alleged that as of the mid-1990s, the agency employed no speakers of Dari or Pashto, the languages common to Afghanistan, and had no operatives in the region. But in August, the CIA announced the graduation of its largest class of new officers ever. Three-quarters of the new recruits are fluent in foreign languages.

The U.S. intelligence apparatus needs to assemble an orchestra of virtuosos in short order, reform advocates say. But all of them needn't be born with powers of intuition or spend their lives in foreign countries. One senior thinker on Islamist terrorism who works in U.S. intelligence strove to see the world through his adversaries' eyes without living among them.

SEEING AND BELIEVING

"Anyone who worked [on] Afghanistan [from the mid-1980s to the early 1990s] had an acquaintance with the name Osama bin Laden," says a 22-year veteran intelligence analyst who penned a provocative study of the terrorist leader. The analyst calls himself "Anonymous." He still is employed by a U.S. intelligence agency. To get permission to publish his book on fundamentalists, Through Our Enemies' Eyes (Brassey's, 2002), he agreed to keep his name and the agency's out of the news media.

When Anonymous started covering Islamist terrorism in 1985, bin Laden's name came up frequently. But analysts were convinced the Saudi millionaire was the financier of terrorist operations, not their mastermind or the inspiration behind them.

Why? Analysts thought Saudis rarely performed their own labor, Anonymous says, a commonly held assessment of Arab social structures. Saudis paid others to clean their homes and pump their oil. Bin Laden supported terrorists' goals, but would never take up their struggle with his own hands. Or so went the theory, which persisted well into the 1990s, Anonymous says.

Anonymous was unconvinced, so he started reading the views bin Laden had publicly expressed, which were carried in some English language papers abroad and by foreign broadcasting services. In 1996, Anonymous changed his perception of how bin Laden viewed the United States.

In August 1996, Anonymous read bin Laden's first declaration of war on America, the "Ladenese Epistle," printed in the London-based newspaper Al Quds Al Arabi. "The people of Islam have suffered from aggression, iniquity and injustice imposed on them by the Zionist-Crusaders alliance," bin Laden wrote. "Greatest of all these aggressions . . . is the occupation of the land of the two Holy Places [Saudi Arabia]-the foundation of the house of Islam." Bin Laden said a "blessed awakening" was "sweeping" the Islamic world, and announced that he would "study the means that we could follow . . . to return to the people their own rights."

Anonymous noticed recurring themes in bin Laden's writing: decrying the occupation of a sovereign land by a foreign power; listing grievances and the wrongs committed by the occupier; and, ultimately, calling for rebellion to throw the invader out. The strains sounded familiar to Anonymous, a student of American history. "It rang so close to our Declaration of Independence," he says.

The fervor and passion of bin Laden's writing frightened him, driving out the financier model of bin Laden. Replacing it, in Anonymous' analysis, were the increasingly common film and photo images of bin Laden carrying guns and eating, drinking and dwelling with non-Saudis. Bin Laden had vowed vengeance on the United States in plain, deliberative language. Anonymous took him at his word.

Anonymous' superiors didn't see things his way. The "mad mullah" model persisted: Fanatics living in caves, with their dirty beards and primitive military technology, never could be a credible threat to the United States. Intelligence agencies still took a Cold War approach to terrorism: for every terrorist group, they sought a state sponsor. Few could accept that bin Laden's group was independent, even though they could see it was self-financed.

"We underestimated bin Laden and what he was up to," Anonymous says. Intelligence agencies failed to grasp that bin Laden was building a global terrorist network of sophisticated, well-financed attackers. The agencies were effectively hobbled in their attempts to predict bin Laden's next moves because they had dismissed and misunderstood him.

But in June 1999, agency officials asked Anonymous, who had by then studied fundamentalism for more than a decade, to write a primer for new analysts about Islamist terrorists. A year later, Anonymous proposed his radical treatise be published as a book. What's more, he planned to support Harvard professor Samuel Huntington's thesis that a "clash of civilizations," not ideology or economic disparity, would frame the wars of the future. Huntington-who had been assailed as a bigot by Muslim intellectuals when his work was published in 1993-was a pariah within the agency, Anonymous says, deemed too offensive to America's Middle Eastern allies. Ultimately, agency officials refused to support publication of Anonymous' book.

He submitted a manuscript to a publisher anyway, and was able to get the agency's approval by agreeing to delete his name and that of his employer.

Anonymous' assertions could be discounted as benefiting from hindsight. The book was published after the Sept. 11 attacks, now attributed to bin Laden's al Qaeda. But the manuscript was ready for publication three months before the attacks took place. "The vagaries of the manuscript-to-book process," Anonymous writes in his preface, "found it unpublished on 11 September 2001.

"But while the events of 11 September and after have been dramatic, bloody, and varied," he continues, "they have not fundamentally altered the story I intended to tell." In an earlier passage the author warns, "The United States has never had an enemy who has more clearly, calmly and articulately expressed his hatred for America and his intention to destroy our country by war or die trying."

Reactions to Anonymous' analysis have been mixed. One former CIA chief said the analogies to American revolutionary history made him want to "retch." And although Anonymous says his managers at the agency have said "zero" to him about the book, they apparently no longer want his analysis of Islamist terrorism. "I am no longer working on these issues," he says.

Anonymous and Peters represent varied strains of analysis. Yet each analyst arrived at the same conclusion: Osama bin Laden and his style of non-state sponsored, global terrorism posed a real and imminent threat.

So, why did the intelligence bureaucracy fail to take such analysis seriously? The skeptical and self-protecting nature of the system likely played a role. Also, human sources of intelligence were lacking. The July 24 report of the congressional inquiry into intelligence lapses preceding the Sept. 11 attacks found that the lack of sources within terrorist circles and operators on the ground in Central Asian nations compounded intelligence agencies' lack of understanding.

In addition to those problems, the agencies have endeavored to speak with a single voice, as an "intelligence community." Critics say the agencies have become victims of "groupthink," a phenomenon that can inhibit debate and prevent unorthodox ideas from being reported forcefully to senior decision-makers. The condition is said to be especially pronounced when representatives from the intelligence agencies come together to write national intelligence estimates, reports on a particular topic or area that are delivered to administration leaders and lawmakers. Groupthink has diminished the quality of this work, critics say, and U.S. intelligence has suffered as a result.

GROUPTHINK

The way groupthink dilutes or discards maverick analysis from national intelligence estimates finds an analogy in events leading to the explosion of the space shuttle Challenger in 1986. In a series of meetings with colleagues and managers during the weeks preceding the shuttle's scheduled launch in late January, some NASA shuttle engineers voiced concern that cold weather could endanger the shuttle's O-rings, the rubber seals between segments of the two solid rocket boosters.

The engineers used data compiled over several years to show that cold temperatures might affect the O-rings' stability. However, others in the launch group rejected their concerns as overblown. Groupthink took over, and the majority view that it was safe to launch the shuttle in cold weather overcame, and eventually consumed, the dissenting opinion.

The majority pointed out that Challenger and other shuttles in the fleet had been launched under cold conditions before without incident. NASA had postponed Challenger's launch four times, and pressure for success was mounting. Among the crew was Concord, N.H., high school teacher Christa McAuliffe, who was to be America's first teacher in space. The Challenger mission was being watched on TV by millions across the country.

Once the O-ring dissenters realized the group was against them, they capitulated, conceding that the shuttle might be safely launched in colder temperatures than they'd originally believed.

Challenger was destroyed when a rigid O-ring failed 73 seconds after liftoff on Jan. 28, 1986. Months after the disaster, NASA determined that the cold temperatures on the launch pad that morning had helped cause the O-ring to crack.

The pre-launch meetings, and the groupthink that prevailed, are chronicled in exhaustive detail by sociologist Diane Vaughan in The Challenger Launch Decision (University of Chicago Press, 1996). Laurie Broedling, the former NASA associate administrator and organizational psychologist, calls the book "one of the most awesome pieces of scholarship I've ever seen," and believes its findings about how groupthink taints decision-making apply at all government agencies, including intelligence agencies.

Broedling, who came to NASA in 1992, when Challenger accident inquiries still were ongoing, says the O-ring engineers were brought back into the majority fold because of "pressure to be permissive." It's a classic groupthink scenario, she says. A dissenter speaks up two or three times, but no one listens, Broedling explains. By the fourth attempt, the meeting has dragged on too late. People start fidgeting. Someone looks at his watch. The group wants to adjourn. And as the dissenter prepares to speak again, he may be met with a fixed glare from a superior, she says. "There's a point at which the cues are, 'You've said it, you've done as much as you can,' and [the dissenter thinks], 'If I say it again, I'm going to be fired or ostracized-there's going to be a penalty here.'"

Groupthink often hijacks the crafting of national intelligence estimates. The officials writing such reports know each other, and some have developed strong bonds, while others carry deep grudges, says Ron Marks, a 16-year CIA veteran who served as an agency liaison to the Senate. As people argue over whether a piece of intelligence should be included in a report, the members take sides, and groupthink emerges as way to achieve unanimity. The result, says Marks, is that truly controversial analysis, which often needs to be part of the estimate, is left out.

Irving Janis, the Yale University social psychologist who coined the term "groupthink" in the early 1970s, found it appeared most often among insular organizations trying to avoid disharmony. The more a group defines itself as a unit, "the greater is the danger that independent critical thinking will be replaced by groupthink," he wrote in his book, Victims of Groupthink, published in 1972 by Houghton Mifflin. The book says the members "wish to preserve the harmony of the group, which inclines them to avoid creating any discordant arguments or schisms." Janis found groupthink affected U.S. decision-makers during the invasion of Cuba at the Bay of Pigs in 1961 and in the events preceding the Japanese attack on Pearl Harbor in 1941.

Intelligence agencies are fraternal and clubby by nature. But in addition, agency employees often adhere to an "us-versus-them" mentality when relating to the policy-makers who use their product, Marks says. The analysts suffer from a "stupid customer syndrome," viewing administration leaders in the White House or heading federal agencies as driven by politics and the expectation that the intelligence will fit their view of the world. At the same time, Marks says, policy-makers fall into a "stupid producer syndrome," thinking the analysts can't know the realities of global strategy, focused as they are on one small part of a big picture. Leaders become exasperated if the analysts can't give them clear, unanimous intelligence reports. This motivates intelligence agency leaders to speak with one voice.

Regardless of where groupthink rears its head, Janis concluded, the outcome is always the same: it leads to inferior decision-making.

But no matter how well or badly it's produced, intelligence ends up in the hands of decision-makers. Some experts believe the intelligence process needs reforming at that level, too.

EXECUTIVE DECISION

The Bush administration's use of intelligence about Iraq's suspected nuclear weapons program, which buttressed the argument for war, underscores the reality that decision-makers use intelligence to suit their needs.

All presidential administrations graze among intelligence reports for the juiciest pieces, says Peters. A senior intelligence staffer in the White House drug czar's office during the Clinton administration, Peters says Clinton clung to reports and analyses that jibed with his interpretations of world events. Marks, the former congressional liaison, served three administrations and concurs that selective use of intelligence is a common practice.

But the recent flap over Iraq's nuclear weapons program also shows how groupthink can overtake intelligence customers. In a New York Post column on July 7, Peters asked, "Was the intelligence community able to pinpoint [weapons of mass destruction] stockpiles with absolutely certainty? Nope," he answered. But if one "weighed the available evidence, in bits and pieces, among like-minded advisers, it's easy enough to see how the strong possibility that Saddam had active WMD programs became first a probability, then a certainty."

Another former intelligence official, who asked not to be identified, agrees and says overreliance on intelligence is misguided. "Intelligence at best should be the spice in the soup, not the soup itself," he says. But "I'm not so sure you don't have people [in the White House] that trust the intelligence too much." Some Bush administration officials seize on pieces of intelligence to prove their point because they think the intelligence agencies aren't aggressive enough in their analysis, the official says. "[They] think intelligence is too milquetoast."

With respect to Iraq, the intelligence producers and their customers are locked in a vicious cycle. The White House reportedly leaned on CIA analysts to make bold conclusions. But the intelligence agencies, reluctant to be radical, tried to qualify their estimate of Iraq's suspected program. Senior White House officials inserted intelligence on Iraq's alleged efforts to buy uranium from Africa in the 2003 State of the Union address after CIA Director George Tenet had warned the White House not to do so. The agency had determined at least one uranium report was bogus.

When the Iraq intelligence estimate was released in October 2002, it distorted the doubts some analysts had raised during its creation about the true nature of Iraq's nuclear aims, according to a former State Department intelligence official. Greg Thielmann, whose team questioned whether aluminum tubes Iraq had tried to purchase were intended to help build a nuclear bomb, told National Public Radio in July that the State Department analysts who contributed to the intelligence estimate "became convinced that it was not likely that this was intended for the nuclear weapons program. . . . We decided that it was nearly certain that this was for artillery rockets."

However, the finished estimate said "most" analysts believed the tubes were for a nuclear program and that "some" believed they were for another purpose, Thielmann said. To a layman, it might look as if the agencies had reached consensus, he said. But the reference in the estimate was "not a true statement. There was no such consensus."

President Bush tied the uranium tubes to an Iraqi nuclear weapons program in his State of the Union address, adding that Hussein "clearly has much to hide."

Richard Kerr, a former CIA deputy director leading an agency review of the Iraq analysis, told NPR that intelligence often is modified as it moves from producers to policy-makers. "Policy-makers . . . trying to present a policy, have a hard time dealing with all the nuances of intelligence," he said. "Intelligence analysts," he continued, "always want to put a bunch of qualifiers in. If you're a policy-maker, you're trying to convey an idea, it is very hard to do that if you have all the qualifiers. . . . Did someone kind of strip it down a little bit? You bet."

But, says the former intelligence official, "when you do that, you've really opened yourself up to some problems, which is that some of this stuff is just going to be goddamned wrong."

In July, the White House released a portion of the Iraq estimate that said intelligence agencies concluded Iraq "probably will have a nuclear weapon within this decade" if it pursued a weapons program unchecked. But the report also included a footnote-a rarely used means of formal dissent-from the State Department's intelligence agency that referred to Iraq's attempt to buy uranium tubes. "Such efforts are not clearly linked to a nuclear end use," the footnote said. It added that "the claims of Iraqi pursuit of natural uranium in Africa are, in [the agency's] assessment, highly dubious."

Asked by reporters at a news conference in July whether the president was aware of State's dissent from the group, a senior White House official said Bush hadn't read the full 90-page report, adding, "The president of the United States is not a fact checker."

ALL TOO HUMAN

The legacy of the Iraq controversy may be to expose the intelligence process for what it is: human, and therefore fallible. Yet two years after the Sept. 11 attacks, some question how the agencies could have missed so many signals that, in hindsight, seem to have painted a clear picture of an imminent strike.

Intelligence is murky, because of how it's collected-from informants, intercepted phone calls, satellite images-and how it's analyzed and used. Analysis is a judgment call. And because it is a human art, it's prone to failure.

To the person, the current and former officials interviewed for this story say that even if they had known all the clues about the Sept. 11 attacks that were discovered after the fact, they couldn't have connected the dots to prevent the events. Peters, who'd written in the mid-1990s that lower Manhattan was an obvious terrorist target, says the best analysts couldn't have anticipated the al Qaeda attack's "sophistication, the elegance, the precision with which they pulled it off. Even had our warnings been taken seriously, there's a good chance [the attack] would not have been prevented."

Government officials and the public may be unwilling to accept that. Nevertheless, the best chance at reforming intelligence for a new era may be to rely on nurturing independent minds, and reversing the tendency toward groupthink and manipulation, even at the risk of some bad or even missed calls.

Cataclysmic events have precipitated calls for "intelligence reform" in the past. Indeed, the CIA was created after the attack on Pearl Harbor. Its mission was to ensure America's adversaries never again caught the country by surprise.

Yet intelligence never has been truly reformed. The natural reaction of bureaucracy is to fend off criticism by adding layers. The Bush administration's boldest solution to counterterrorism has been to add the Terrorist Threat Integration Center. Established in May, the center is the new home of terrorism analysts from across the government. It is overseen by the CIA and housed at the agency's headquarters-"the campus"-in Langley, Va.

If beautiful minds, more than the rearrangement of agencies, are the best hope for true reform, then decision-makers may not have to look very far to find them. Says Peters: "There are more talented people in the intelligence community than the intelligence community realizes."