More Art than Science: Intelligence and Technical Topics

April 12, 2016

For special access to experts and other members of the national security community, check out the new War on the Rocks membership.

In intelligence circles, gaining access to a senior policymaker for a 20-minute “round table” to communicate the finer details of an intelligence product is an incredible luxury. Written intelligence reports today are often meant to be absorbed in a matter of minutes, with little time for a follow-up briefing. Intelligence products add unique value since they are objective and based on sensitive sources. However, from the perspective of a busy policymaker reading a report in between meetings, intelligence products can also come across as simplistic or confusing. Incorporating complex scientific and technical information can complicate intelligence products even more, depending on the experience of the consumer. The contentious 2002 National Intelligence Estimate (NIE) on Iraq’s Continuing Programs for Weapons of Mass Destruction is an important example of an intelligence report that relied on conveying scientific information. That report attempted to convey a highly technical assessment of Iraq’s suspected weapons programs in the space of a few sentences, while still retaining confidence levels and sourcing.

Scientific and technical intelligence analysts thus face the great challenge of quickly, effectively, and clearly conveying information to policymakers. An example of meeting this challenge is aptly illustrated in the book Most Secret War, one intelligence officer’s account of a high-level meeting that occurred over 75 years ago.

In September 1940, the German Luftwaffe began nighttime bombing raids against British cities. The German bomber pilots used a new type of radio navigation that improved their precision and lethality. A young British intelligence officer named R.V. Jones was instrumental in discovering and explaining the scientific technique used by the Luftwaffe:

[The meeting] must have already been in progress for some 25 minutes when I arrived, and I listened for a time while some of those around the table made comments which suggested that they had not fully grasped the situation; only then did Churchill address a question to me on some point of detail. Instead of dealing with it, I said, “Would it help, sir, if I told you the story right from the start?” Churchill seemed somewhat taken aback, but after a moment’s hesitation said, “Well, yes it would!” And so I told him the story. … Churchill himself recorded that I spoke for some twenty minutes, which is quite a time to have the Prime Minister listening at the height of the greatest crisis that had ever confronted the country.

R.V. Jones realized that if the intelligence consumer does not fundamentally understand the scientific concepts and implications involved in the issue, the policy discussion can flounder without complementary methods of describing the information.

A 2009 article in The Economist showed that, worldwide, most politicians come from a legal, business, diplomatic, or military profession. In contrast, only about 7 percent of them come from an engineering, medical, or academic background. Even if any sort of scientific background were more common in the national security community, the specific nature of foreign policy problems makes it unlikely that a given consumer would immediately understand the science behind each issue. For example, an aeronautical engineer would probably comprehend much of the detail related to a missile launch, but that individual might not have the background to discuss the intricacies of a biological weapons program. Science naturally involves the work of specialists, but policymakers make decisions as generalists, especially when they work at the strategic level.

Because policymakers are most often generalists, scientific information in intelligence must be “translated” for the consumer into more accessible language. Scientific language is often unique because intelligence on other topics, such as political or military issues, often falls within the experience of the policymakers and does not incorporate the same jargon and style that scientists cultivate.

Moreover, cultural norms in the scientific community often separate it from the non-scientific community and affect communication. For example, in a December 2014 conference on communicating science topics in the defense community, one panel member noted, “… many researchers seem incapable of explaining what they do to nontechnical audiences, and often fail to recognize the source of audience incomprehension.” Many scientists are so immersed in their field that they have lost the ability to connect with non-experts.

Perhaps the most well-known stylistic difference between everyday communication and a scientific document is the use of jargon, or specialized language, in the latter. Scientists are specifically trained to use specialized language that does not appear in common usage, but which helps them communicate more precisely with other scientists. The author of a recent academic paper summarized the problem: “… the process of becoming a scientist inevitably involves learning to talk and write science according to the norms of the scientific community. These would include generalizing and abstracting rather than building on examples, stories, and anecdotes, and using accurate descriptions rather than analogical approaches. …”

The field of science communication is the study of the problems and potential solutions in communication between scientists and non-scientists. Although it is a recent academic field, some research on best practices for written communication has emerged and presents a new opportunity for the intelligence community. Specifically, harnessing lessons learned from research on science communication could aid intelligence officers that research technical topics, as well as their reviewers and editors, in communicating with senior policymakers. Science communication research offers methods of conveying concepts through linguistic techniques like figurative language and stories.

These methods could be applied to novel intelligence product lines while still meeting the analytic standards described in Intelligence Community Directive 203. Products like the National Intelligence Estimates are bound by years of tradition and legal and bureaucratic requirements that dictate particular formats and styles for their language. These requirements generally succeed in maximizing the precision of language employed — every word is carefully selected and placed.

However, alternative, complementary products should be considered, especially given that different policymakers may consume information in different ways. While it is important to clarify the Intelligence Community’s position, it is also worthwhile to produce products simply to inform the policy debate, as R.V. Jones did when he spent twenty minutes simply “telling the story,” rather than answering the specific question. Parallel examples can be found in unexpected places, such as museum exhibits and public health campaigns that convey scientific-themed information to diverse audiences. These alternatives in science communication may provide new ways to convey scientific and technical intelligence. High-level policy meetings bring together individuals with unique backgrounds and perspectives. Better intelligence levels the playing field, allowing for a more grounded discussion less influenced by public speculation or misinformation.

Conversely, intelligence professionals would do well to also remember the words of one journalist who stated: “the precision of the layman is an accuracy of impression rather than an accuracy of specific fact.” In policymaking, both types of accuracy are often required, but accuracy of impression is more easily lost. Accuracy of impression might be improved with different intelligence products not bound solely by the requirements of National Intelligence Estimates, or by the trappings of scientists.

R.V. Jones ultimately helped address one of the most important Analytic Standards that exists — Tradecraft Standard 5. He demonstrated customer relevance and addressed implications by more effectively conveying technical concepts and providing context. His timely insight and methods helped save his country by allowing Churchill to craft a well-informed response to the enemy attacks. Today’s intelligence community needs to learn from all of these examples and apply them in new ways if we’re to do the same.


Dr. Brian Holmes is a faculty member at the National Intelligence University in Washington, DC. Max Greenlee is a National Intelligence University alumni and an employee at the National Geospatial-Intelligence Agency. The views expressed in this article are theirs alone and do not imply endorsement by the Defense Intelligence Agency, the National Geospatial-Intelligence Agency, the Department of Defense, or the U.S. government.

We have retired our comments section, but if you want to talk to other members of the natsec community about War on the Rocks articles, the War Hall is the place for you. Check out our membership at!

2 thoughts on “More Art than Science: Intelligence and Technical Topics

  1. There is a similar need from the scientific research point of view. Science communication is imperative to getting funding in order to produce more science. Every successful scientist needs to “sell” their research to someone not quite in their field. Generally, the larger the amount of money you are trying to acquire, the more general your explanation must be. One’s level of research explanation would for instance progressively generalize within academic institutions from immediate research peers, to institute director, to dean of faculty, to president of the university, to the state government, to the federal government. The biggest financial success comes from making the most general explanation to the person or agency highest up the chain. So there is a similar incentive in academia to provide different product levels of communication.

  2. Specifically, it seems that your reports could benefit from a non-sequential, layered approach to information reporting. Reports with something similar to embedded hyperlinks, that expand and generate more detail, would allow you to present a simple, collapsed 5 minute report, but with the opportunity for the more curious or technically inclined consumer to gain more insight. The consumer could “click” on sentences to expand each generalized point to gain an appreciation for the detailed analysis that produced it. The format of such a report structure could be made in a secure, non-commercial, self-contained, simple format. To be secure, one would want it self-contained, and not linked to external code. If there are constraints that the report must also be printable in a sequential text format, this could also be accommodated. I have some ideas for such structures if you are interested. It seems like an interesting project to devise such a way of layered information dissemination.