05 December 2005

Source Reliability

I am rather offended by the direction that the "extraordinary rendition"/"rigorous interrogation" debate has gone, primarily because it is based on unreliable sources. I don't just mean those being interrogated, who are (by their nature) unreliable sources. I mean those contributing to the debate itself. Go ahead—name even one individual on the "pro-torture" side of the debate who has publicly contributed to the actual policy debate raging in Washington (and elsewhere) who has hands-on experience interrogating non-American intelligence sources. John Ashcroft and Alberto Gonzales don't; Jay Bybee and John Yoo don't; and so on. I cannot name a single Administration official, or even not-in-the-government shill, who has such experience—even second-hand, in the sense of directly supervising the interrogation. My knowledge of this is not encyclopedic; I'd welcome documentation demonstrating otherwise.1

Of course, interrogation of foreign information sources is very far outside of the American experience (which is part of my point). Let's look at it in the context of something more familiar to Americans (and, for that matter, blawggers and policy wonks—very few of whom have ever gotten their hands dirty with anything other than ink stains). Assume for the moment that you are the newly appointed CEO of a moderately large manufacturing corporation. The top folder in your in box concerns purchase of tools and manuals for the machines on your production line (we'll assume, for the moment, that you're not in a position to change the machines themselves). Your objective as CEO is to get the maximum output from your production cycle that you can, consistent with whatever quality standards you have set. The folder includes two memos.

  • One of the two memos was written by a 25-year-old kid with an MBA. It includes lots of fancy charts and graphs, and has extensive appendices filled with intimidating formulae, offprints from Business Week and Forbes, and copies of the sales brochures from the two companies that make tools compatible with your corporation's current machinery.
  • The other memo was written by a curmudgeonly shop foreman who worked on the two-generations-ago machines, the last generation of machines, and the current generation of machines, and has a total of thirty years' experience. According to your executive assistant, this is the guy that the tool companies go to when they have questions about how well the machines work.

It is entirely possible that the two memos make the same recommendation. It is also entirely possible that the whiz kid is right and the shop foreman is wrong. The one thing that you, as a CEO, are not going to do, though, is make your decision solely upon the basis of the whiz kid's memo. If he's right and the shop foreman is wrong, as the CEO you'll need to be able to articulate why to your stockholders, whether or not something goes wrong.2 At minimum, as a responsible CEO you need to feel like you have adequate information yourself to do your job well.

This is the same problem that I saw all too many times in the military. Due to various peculiarities in the military hierarchy, a fair number of units end up getting commanded by individuals with little or no background in the unit's mission or capabilities. In theory, their subordinate commanders should catch any mistakes before they get serious. The problem comes when the command element all has a common background… and it doesn't match the unit's. Consider the fighter pilot given command of the supply squadron; the naval aviator given command of a carrier; the career artillery officer given command of a military police battalion. I have personally observed the results of each of these decisions (don't ask, I won't tell); in each instance, the "outsider" with no context for his decisions screwed up the unit in question for years to come.

That's what I see going on in the "torture" debate. The people we're not hearing from are those who would be able to evaluate "torture" against other methods of gathering intelligence, considering accuracy, timeliness, and the effect on other sources of intelligence. It's not just that—as is proper—they're keeping quiet because they're too busy doing their jobs (and not interested in undermining their own efforts). It's that they're not even being asked.


  1. Admittedly, this is going to be rather difficult, based on the typical NDA to which one who has such experience is bound. Some of the NDAs are themselves classified…
  2. Unless your corporation is an imperialist regime that at best gives lip service to shareholder voices, rights, and interests, not to mention the interests of its customers. <SARCASM> I can't think of any corporations that operate like that. Well, OK, Disney. And RJ Reynolds. And… </SARCASM>