By John A. Irvin (Based on comments made during The Botstiber Institute for Austrian-American Studies panel discussion “Espionage and Intelligence in the 21st Century.”)

World War II, and especially the subsequent Cold War, woke America up to the importance of continuous, institutionalized intelligence collection.  Initially we followed the example of most other nations and focused on Human Intelligence (HUMINT) – recruiting and handling human assets.  This wasn’t just tradition, but an acknowledgement of the limitations of technology.  While the United States certainly engaged in photo reconnaissance and intercepted communications, at the time the best way to get the secrets our country needed for national defense was to recruit individual foreign citizens with access.

As the Cold War progressed, the US developed a significant technological advantage over our Soviet-bloc adversaries.  In the military this was expressed in weapons that were effective enough to more or less balance the Soviet’s advantage in sheer numbers.  In the world of intelligence collection this was expressed in machines that gave us an advantage in technical collection like Signals Intelligence (SIGINT) and imagery.  In both the military and intelligence collection, it was only natural that we would lead with our strength.

This certainly isn’t to say the US abandoned HUMINT.  Open-source history clearly shows otherwise.  However, more and more focus and resources were gradually shifted toward technical collection.  This makes sense for a number of reasons.  First, it flowed from the US advantage in all forms of science and technology.  Second, technology derives from fields like mathematics, physics, and engineering; so the product of technical collection is more or less tangible.  You may misinterpret it, but it is what it is.  It’s objective.

On the other hand, HUMINT derives from psychology; the personal relationship an asset has with his or her handler, their conscious and unconscious motivation and bias, their world view, their ego.  So the product is more subjective and contextual.  From the asset to the handler, from the analyst to the policy-maker, HUMINT passes through a series of individual minds, each with their own distinct view of how things “really” are.  Nothing is a better predictor of plans and intentions, but by the same token nothing is as prone to cognitive bias at all levels.

There are basically two ways humans get what we need – hunting and gathering – and we’ve been practicing both for millennia.  When you hunt, you go out deliberately seeking a specific thing.  Whether or not you come back with it depends to a great degree on your skill, training, planning, and luck.  When you hunt you’re more likely to find exactly what it is you’re looking for, to make the big kill.  But you’re also more likely to come back empty-handed.

When you gather you simply take as much of what’s available and then sort through it to find what you’re looking for.  In some cases you do literally cast out a wide net over and over again.  You’re less likely to find exactly what you’re looking for, but you’re also less likely to come up empty-handed.  If you work hard enough, or you’re lucky, you’ll eventually come up with enough to satisfy your needs.

Of the two, hunting requires more training and experience for a given individual to be successful at it.  Gathering doesn’t require as much training and experience, but it does require a great deal of effort.  It’s extremely labor intensive.  In the modern era we’ve been able to steadily reduce human labor through our invention of ever more complex and ingenious machines.  Once the machine or the software program is in place it makes gathering the easier of the two practices (although gathering too much introduces its own problems).

Technical collection can give you vast amounts of information, more than you can get from any single HUMINT asset.  However, all action originates in the minds of individuals, so unless individuals deliberately placed their thoughts in some medium that is accessible by technical collection, there’s simply no way of knowing their plans and intentions (especially when they’re deliberately trying to keep them secret) except through a personal relationship in which an asset is willing to tell you what he or she is thinking.

Where we are today is, ironically, best summed up by NSA-leaker Edward Snowden.  In his most recent interview he said, “It’s no secret that the US tends to get more and better intelligence out of computers nowadays than they do out of people.”  That’s like saying thirty years ago that the US tends to get more and better intelligence out of filing cabinets than from people.  Snowden’s statement demonstrates an all-too-frequent mindset in which the focus shifts from the human to the technical, where the means take on more importance than the ends.

The computer or the filing cabinet doesn’t generate intelligence, it’s just where information is stored, some small part of which may actually be intelligence.  A computer file or voice recording is just the tangible record of a thought expressed.  Before it ever made its way to being a technical collection product, it sprung from the mind of an individual.  Some of those thoughts, perhaps the most important ones, never get typed into the computer or mentioned over the cell phone.

Espionage is, they say, the second oldest profession, which puts it in questionable company but attests to the fact that it predates computers and satellites and just about everything we associate with intelligence collection today.  It’s a uniquely human activity, and while we continue to develop increasingly more complex machines to do it with (as we well should in a dangerous world), we should never ignore the fact that we’re the ghost in the machine.

Machines don’t keep secrets, people do.  Our analytic decisions and the foreign policy consequences that derive from them don’t come from machines, they come from something even more complex – unique, individual human minds, each one the product of a lifetime of experience, emotion, choices, beliefs, and biases, not all of which the individual may even be consciously aware of.

Moreover, our means of technical collection are moving out of the hands of governments and into those of private citizens.  Recent events would suggest that you as a private citizen should probably be more concerned that your aggrieved lover might be recording your conversation rather than the NSA.  Or that your potentially compromising picture is being taken by a stranger with a cell phone rather than an overhead satellite.

Michael Warner makes this point very well in his recent book The Rise and Fall of Intelligence.  Technology advances and becomes ubiquitous at a rate faster than our human ability to fully adapt to it.  The brain evolves at a much slower rate, to the point where the human mind is the one part of the equation that remains virtually constant.  It might be time that we dedicated as much attention to the human side of intelligence collection as we do to the technical side…as much to the psychology of espionage as we do to the technology.