By Erik Kleinsmith, Staff, Intelligence Studies, American Military University
[This article originally appeared at In Public Safety]
Intelligence analysts like to solve problems. We really do. We like to take various points of information and generate a deeper understanding of a particular subject so that we can better answer specific questions, mainly about things that are going to happen in the future. But in our quest to find solutions and answer questions, our solutions quite often end up causing additional problems.
One question we never seem to solve is about our own profession—whether intelligence is an art or a science. Opinions widely differ. In professional circles, newsgroups, and forums, the question of whether intelligence analysis is an art or a science is a perennial topic of discussion and disagreement. To liberally paraphrase Mark Lowenthal, a former Assistant Director of Central Intelligence for Analysis and Production at the CIA, in an informal speech a few years ago, “It’s an art. That settles it. Quit talking about it.” As a friend and colleague, I consider Mark one of the founding fathers of modern intelligence analysis, and yet, here we are still arguing one side or another.
After 20-plus years of working in intelligence, I’ve come to the conclusion that this question will endure for the ages. Some thinkers on intelligence consider it an art, full of free-flowing ideas, imaginative thinking, and creative inference. Others offer opinions that are more technical in nature, driven by hard data and logical thought processes.
Perhaps the best way is to look at intelligence is as a spectrum with two extremes, instead of a simple either/or question. Every intelligence practitioner is at some place on this spectrum. Most pragmatists (and those who’ve not considered the question) are somewhere in the middle. If you’re not sure where you are, let’s start with a series of qualifiers:
If you view intelligence analysis as:
- Powered by technology
- Structured, with rules
- More deductive in nature
- Quantitative over qualitative information
- Driven by structured queries and logical reasoning
- Made up of hard data sets
- Geared for standardization and collaboration
You most likely will regard it as a science. Men like Sherlock Holmes and Mr. Spock would agree with your perspective. Intelligence has to have structure and consistency. All else is simply throwing darts against a target while blindfolded. Those analysts who lean toward viewing intelligence as a science often operate under the consistent need to show their math and believe the more data, the better. For them, logic and reasoning are the keys to blowing off the doors to fallacies, conjecture, and outright myths. Intelligence has to be explained logically and, like scientific methodology, it attempts to prove or reject a theory. By gilding analysis with scientific structure, it is conducted in a way that can be consistently reproduced under similar conditions.
Conversely, if you view intelligence analysis as:
- Powered by imagination
- Free-flowing, unconstrained
- Inductive over deductive
- Qualitative over quantitative
- Driven by holistic research and critical thinking
- Made up of social/fuzzy problem sets
- Geared for the individual analyst’s creative thinking
Then you most likely regard it as an art. Analysis to you is something akin to playing the violin or any other classical instrument; you don’t become a master until you can capture the emotion of the song in your performance. Mountains of data sets, inputs, and sources of information are so overwhelming to the human mind’s ability to logically reason that it must be augmented by subconscious levels of thought. We create mental models through instinct and emotive logic in order to make sense of most things. While gathering data is important, there comes a point where it’s just noise that you need to sort out in order to find the right tones. To you, intelligence is more along the lines of Picasso and Rembrandt, who made canvas come alive using pigments, or JRR Tolkien, who created an entire new “Middle Earth” going from an invented language to a fantasy masterpiece.
A More Pragmatic Approach to Intelligence
While these two schools of thought on intelligence are distinctly different from each other, they are also on the ends of the analytical spectrum. Rare is the analyst who approaches one of these perspectives as a purist. Most of us are somewhere in between. Perspectives of intelligence analysis are crafted by experience. Different kinds of intelligence analysts tend to lean toward one or the other, depending on the type of analysis they regularly do.
For example, SIGINT analysts, those who specialize in signals, communications, and electronic signatures, lean heavily to the scientific/technical side of intelligence. Their world is one of 1s and 0s, technical collection platforms, and sifting through the never-ending deluge of radio, Internet, and cellular traffic. Their analysis is much more beholden to technology, making their processes more rigid and restrictive, but easier to replicate and standardize. As a result, their analytical methods are tailored to handle massive amounts of technical data. Adversely, analysts of this nature tend not to have the flexibility in their analytical methods to adapt to more human environments.
Logically, these types of analysts view their work as extremely scientific. Those on the extreme end believe that in time, all analysis will be conducted by data-mining tools and complex algorithms. While working in a classified facility, a colleague of mine once proclaimed that the last intelligence analyst has already been born. This is because, in his opinion, computers will replace the need for humans to conduct intelligence analysis. We call people of this perspective software engineers and IT specialists.
Counterintelligence and human intelligence (HUMINT) analysts represent the other side of the art versus science spectrum. To them, intelligence analysis is getting into the human mind and our decision-making processes. Understanding social networks and human interaction is a measure of skill and tradecraft that operates with some basic rules and methods, but can never be quantified or modeled with 100 percent accuracy. There are just too many variables to factor in, so gut feeling, intuition, and logical leaps between datasets are necessary in order to paint a most complete picture. Analysts who lean toward the artistic side of intelligence have processes to be sure, but they are merely guidelines with room to deviate according to the environment and situation. Operate in this way with technical data and your results will be unreliable, unverifiable, and rarely repeatable.
For many of us who learned intelligence analysis through the fast-paced, bottom-line, up-front demands of the tactical environment, analysis is less about art or science than practical necessity. In learning about and working analytical problem sets, most analysts have had little time or reason for stepping back to answer this question. Intelligence analysis was work, no matter if we thought we were using a wrench or a brush.
Just as music is powered by a rhythmic beat (except for jazz, which the structured side of me doesn’t quite understand), dance consists of regimented movements, and painting is controlled by the application of wet paint to a surface, the art of intelligence is powered by scientific facts, data, and processes.
So let’s look at intelligence in a more pragmatic way, as an art that is powered by science:
- Powered by problem sets/requirements
- Tailoring tools and methods with imaginative application
- Applying methodology to better understand and explain complex problems
- Able to find the qualitative aspects of quantity-based data sets
- Geared for both individual and collaborative analysis
- Better able to adjust to fast-paced, fluid environments
- Using a larger and more diverse set of analytical tools and methodologies
Intelligence is an art to settle the ultimate question, but it’s important to view it as being powered by scientific methods, structured data and processes. We rely upon technical means of collection, but the essence of analysis still remains the analyst. New technology is being developed all the time, but nothing will ever replace critical thinking about a subject.
The key to approaching intelligence pragmatically is that your efforts are focused on your customer, their requirements, and the problem sets they throw your way – as opposed to being bound by a process or standard way of doing business. Purely scientific approaches tend to bind us into certain ways of doing business. Purely artistic intelligence tends to get us out of focus. Intel focused on the customer helps to focus more creative or artful analysis by demanding adherence to the rules of reality. Imaginative thinking must have a timeline to adhere to and an eventual answer to questions that your customer is asking. Intelligence for the sake of intelligence is largely a wasted effort.
In the course of your career you will work with analysts from many different walks of life. If you are lucky, you will gain an appreciation for both views and ultimately come to see them as complementary rather than in conflict.
About the Author: Erik Kleinsmith is the Associate Vice President for Strategic Relationships in Intelligence, National & Homeland Security, and Cyber for American Military University. He is a former Army Intelligence Officer and the former portfolio manager for Intelligence & Security Training at Lockheed Martin. Erik is one of the subjects of a book entitled The Watchers by Shane Harris, which covered his work on a program called Able Danger tracking Al Qaeda prior to 9/11. He currently resides in Virginia with his wife and two children.