Blind Zeal as Expertise

Timothy Naftali, a so-called “expert” in the history of intelligence and spying, has no clue what he’s talking about. The following interview is intended to be a discussion attempting to understand the complexities on why the U.S. government would need to skirt around the Foreign Intelligence Surveillance Act, but Naftali’s gushing tone, to say nothing of his lack of nuance in examining the issue in question is baffling in its stupidity. Coming across as a big-time NSA booster on Morning Edition, Naftali let loose the following priapic monomania this morning. Amazingly, he’s an associate professor.

A: If you accept, and I do, that there is the possibility of al Qaeda or its affiiates having cells in this country, how do you monitor these people? If they’re changing their cell phones and if they’re moving from computer to computer. How do you do that?

Q: The PATRIOT Act appeared to address that very problem.

A: No, it didn’t.

Q: Made it possible to give a warrant that will follow an individual from one telephone to ano–

A: But what if you don’t know the individual, Steve? What if you’re looking for patterns of behavior? What if you don’t know the individual’s name?

Q: What if you don’t know the number? How do you follow one person around? From going into Wal-Mart and buying a cell phone?

A: You don’t follow one person around. What you do is you listen to conversations.

Q: You mean you listen to a million random conversations hoping to hear this guy?

A: The White House is saying that it is very careful not to listen to point-to-point conversations in the United States. From one point in the United States to another. But there is a way through data mining to analyze where calls originate and where they go. This is a — basically an attempt to look for patterns, use of words, length of telephone calls, length of email, frequency of these communications, both voice and data, and then to look for suspicious patterns. And how do you define suspicious? I don’t know. But the now Deputy Director of National Intelligence, Michael Hayden, has talked about there being a subtle soft trigger. A computer learns what’s suspicious and then it will act on its own. So what we’re talking about is a higher order — a smarter Google if you will.

* * *

In other words, the hard line being espoused here is a technological miracle for an unspecified and unproven pattern for unspecified profiles. Not even Naftali can determine what patterns might constitute “suspicious” behavior and yet he’s gung-ho to micturate on the Fourth Amendment entirely on speculation. I mean, am I suspicious because I happen to enjoy shopping for groceries at 3 AM? Or because I send emails at odd hours? Or, for that matter, have the courtesy to reply to people with lengthy emails? (Insomniacs, of course, are the most suspicious Americans of all.)

What makes Naftali sound even more like an insouciant fundamentalist is his notion that a self-learning computer that will magically stumble upon the right formula. In the interview, he doesn’t cite a single example why we should subscribe to this methodology, although he offers oblique references to John Poindexter’s TIA project (addressed in this letter).

Make no mistake: this is blind zeal masquerading poorly as analysis. It fails to offer a multilateral take on the subject. It fails to offer a solid benchmark on the current state of intelligence, whether there’s too much spying or not enough. It fails to consider the American concern for privacy or Plamegate or the considerable intelligence that the Bush administration failed to check up on before 9/11, well before the PATRIOT Act. (Could it be that the DOJ, the FBI and the CIA was doing just fine before any of this craziness went down and that the incompetence from top-level administration is the cause? Why is it that nobody bothers to dwell on this question?) It is exactly the kind of dangerous “expertise” that doesn’t even scratch the surface of what’s wrong.

Leave a Reply

Your email address will not be published. Required fields are marked *