Perspectives on the NSA and PRISM: What Dyson Missed

July 28, 2013 Dennis D. McDonald

In NSA: The Decision Problem George Dyson lays out a credible historical view on why he thinks what Snowden revealed about the NSA and PRISM was inevitable. While it’s certainly a pleasure to read a discussion of these issues that is so thoughtful, I think he misses one major theme that helps us understand our current predicament over what to do about Snowden.

Dyson presents useful historical context about how we have always taken advantage of new technology to spy on our adversaries. He also intermingles the relationship between our increasingly sophisticated “machine intelligence” capabilities and the vastly greater volume of communication signals worldwide that can now be monitored and interpreted.

I’m not in complete agreement with his distinction between our wanting to know what our adversaries are saying and wanting to know what they are thinking. This distinction seems to makes sense on the surface but also raises both philosophical and physiological questions that are, in my opinion, difficult to resolve, partly because human memory can also play tricks on us even when “augmented” by the Internet.

For example, if you don’t experience and sense something directly, how can you ever be sure that what gets reported to you via any intermediate medium or channel is completely “accurate”? As humans we only experience and retain a small portion of what surrounds us. For example, our eyes are sensitive to only a small portion of the electromagnetic spectrum. We regularly use a variety of technological systems to extend our grasp throughout all regions of that spectrum. Our doing this is a natural extension of human toolmaking.

It is natural that we use our machines and their constantly improving information processing powers to see over the horizon and — according to Dyson at least — into the “minds” of our adversaries. Still, we have to interpret all this information and make sense of it. This involves using our judgment about what to trust, what to believe, and what to do about what we do come to believe, even in common situations where we lack 100% certainty.

It also means we have to decide how much authority and responsibility we are willing to give to our information gathering and interpretation systems. That’s a problem that is difficult to completely automate.

One thing that Snowden revelations make very clear which I believe Dyson misses in his article is that traditional distinctions between “us” and “our adversaries” are difficult to maintain consistently in a world where communication is instantaneous and different cultures are only a plane hop away. Deriving meaning from communication signals usually requires information about both the source and receiver; that’s basic communication theory. Our desire to know about both source and receiver is at the core of what is so troubling to me about what Snowden has revealed:

We gather information about US citizens. This gets us into trouble with constitution based privacy and surveillance principles.
We gather information about non-US citizens. This gets us into trouble when the country in which they reside is not an official “enemy.”

Of course, anyone familiar with how intelligence gathering operates knows that “friends” will always spy on “friends.” You never know when a country that’s your friend will become your enemy. It’s also possible that some of your friends will be communicating with some of your enemies even if you don’t.

The bottom line: not knowing the capabilities of potential adversaries has always been viewed as a mistake. Now, though, our “adversaries” aren’t as easily defined by national borders. Those who wish us harm can reside in friendly states and vice versa. If you follow this line of reasoning — which I don’t — spying on anyone who has the potential for causing harm or knowing about the potential for harm to U.S. citizens appears to be at the core of some of the most notorious NSA/Snowden revelations.

I’m not excusing the situation but I think it naïve not to take these considerations into account when debating these topics. Such debate is good, especially if it surfaces the complexities of the issues for more people to consider, such as how much leeway we give the Executive branch for targeting adversaries without a legal declaration of war.

I also think that focusing too narrowly on support for a “Free Snowden” movement, for example, would be shortsighted by taking us down a legal road that focuses too narrowly on the question “did he break the law.” Such an overly legalistic view sometimes generates problematicly narrow results, as we saw recently in the Zimmerman murder case in Florida.

If we could focus on details associated with these larger issues, we’ll all be better off, but only if the issues can be debated in public. That means there’s going to have to be a lot more honesty about what’s going on than we’ve seen in the past.

I’ve also seen where the Snowden case is being used as another excuse for more bashing of the United States. I’m not excusing the U.S. at all but I think it’s also naïve to suggest that the U.S. is the only country engaging in PRISM-like cross-border intelligence practices. Failure to it knowledge that does us all a disservice by perpetuating the simplistic “us vs. them” meme.

Related reading: