Spy agencies team up with National Academies Expanded ties include new board, first-ever survey of social sciences and security By Jeffrey Mervis In an unprecedented move, U.S. intelligence agencies are teaming up with the nation’s most prestigious scientific body in a bid to make better use of findings from the country’s leading social and behavioral scientists. The partnership between the Office of the Director of National Intelligence (ODNI) in Tysons Corner, Virginia, and the National Academies of Sciences, Engineering, and Medicine aims to build bridges between communities that historically have either ignored one another or butted heads. The effort includes the creation of a permanent Intelligence Community Studies Board at the academies, which will meet for the first time next week, as well as a first-ever study of how social and behavioral science research might strengthen national security. David Honey, ODNI director of science and technology under Director of National Intelligence James Clapper, says he hopes that the new partnership will help the intelligence community improve how it collects and analyzes information. He and others are eager for help picking out useful and relevant research, as well as grasping where there is a lack of good science. Understanding “the limitations of our knowledge,” says Robert Fein, a national security psychologist in Cambridge, Massachusetts, and a member of the new intelligence board, “will help to protect us against armies of snake oil salesmen.” One area in dire need of better research is figuring out when people are lying, Fein says. After the start of the wars in Afghanistan and Iraq, he notes, intelligence agencies poured money into research on both mechanical-—think polygraphs—-and behavioral-—think interrogations—-methods of detecting deception. But the results were disappointing, recalls Fein, who led a 2006 report on interrogation techniques for the director of national intelligence. “Researchers overpromised,” he says, “and there were few useful results after millions of dollars were spent.” Even research that generated solid results had serious flaws, he adds. “For example, none of the studies [of deception] involved people who didn’t speak English,” Fein notes, making them of questionable value for use in many current hot spots around the globe. Last week, Honey harvested the first fruits of the fledgling collaboration at a 2-day public summit in Washington, D.C., designed to feed ideas into the upcoming survey of social and behavioral science. Prominent researchers presented talks on everything from decision-making under stress to how social media fuel conspiracy theories. In addition, a panel of intelligence analysts offered a rare glimpse into their classified world and how scientific results need to be tweaked before they can be applied to intelligence and policy. “Scholars like to say, ‘In general, X is the case.’ But as practitioners, we are asked to respond to a specific situation,” explained Charles Gaukel of the National Intelligence Council in Washington, D.C., which consists of senior officials from each of the 16 intelligence agencies across the government. Gaukel also debunked the popular notion that intelligence analysts try to forecast global events. “Our role is not to make predictions,” he said. “Rather, we try to give policy-makers a sense of what’s out there, and how the enemy is likely to respond.” The new survey won’t follow the usual blueprint for so-called decadal surveys, which look ahead 10 years. Traditionally, decadal studies help a particular discipline, such as astrophysics or geoscience, set priorities among competing facilities and projects. They may also recommend how federal agencies can fill gaps and maximize their research portfolios. In contrast, the new study won’t seek to balance competing demands for scientific facilities or examine existing research portfolios at the intelligence agencies. Instead, Honey hopes it will identify current and future research areas that might be useful for national security. Speaking at the summit, Fein suggested that those on the decadal survey might want to convene a panel of intelligence practitioners who would propose areas “that might benefit from the relevant perspectives, data, and knowledge” of social and behavioral scientists. “I can’t forecast what we will learn” from the study, Honey says. “But the history of decadal studies shows their value in pointing to where the research community is headed.” The survey is expected to take 2 years, and more than 300 people have been nominated for the panel’s 18 slots. Despite that strong show of interest, some scientists who support the effort worry that getting involved could harm their research because of public unease about the spy agencies’ activities. “I’m a citizen scientist, and I think this collaboration is great,” says Paul Glimcher, a neuroeconomist at New York University in New York City who spoke at the summit about the Kavli HUMAN project, a deep dive into the behavior and characteristics of 10,000 New York City residents that he leads. But he says his team has “promised our subjects that the data will never be shared with the government. And I’m concerned they might react negatively to my being involved in a study funded by U.S. intelligence agencies.” Some observers see the new intelligence board as a successor to an ODNI science advisory panel that Clapper abolished after becoming the nation’s top spy in 2010. “It seems like at least an admission that the abandonment of [that advisory board] was a mistake and that there is an important role for independent advice from academics to the intelligence community,” says Steven Aftergood, who leads the government secrecy project at the Federation of American Scientists in Washington, D.C. Gaukel may have inadvertently given the pending decadal survey its marching orders during his summit talk on what intelligence analysts do. “We’re looking for truth,” he said. “But we’re particularly looking for truth that works.” Science 14 OCTOBER 2016 • VOL 354 ISSUE 6309 DOI: 10.1126/science.354.6309.155 http://scihub22266oqcxt.onion/http://science.sciencemag.org/content/354/6309...