Along the Tracks

Friday, May 17, 2002

Connecting the dots

The uproar over “revelations” that Bush was warned about a hijacking threat last August, and the mostly unspoken but clearly present subtext - that he could have stopped Sept. 11 from happening - is such ridiculously partisan finger-pointing that I can hardly believe it is occurring. In particular, I want to focus on the “connect the dots” criticism: That is, why didn’t the White House “connect the dots” and see the attacks coming?

For you and I and the reporters asking the question, the connections between the following dots seem obvious on their face: a) an FBI agent is concerned that al Qaeda sympathizers are taking flight training in the U.S.; b) an al Qaeda member has been arrested in Minnesota, where he raised suspicion by taking flight lessons and saying he didn’t need to know how to land; c) overseas intelligence information, several years old, indicated al Qaeda hijackers were hoping to crash planes into the Eiffel Tower and the CIA headquarters; d) the CIA had information al Qaeda was plotting airplane hijackings. America + flight schools + hijacking + didn’t need to know how to land = September 11! But there are a few other details which are being conveniently ignored in this shallow assessment.

A) The volume of information. The dots listed above are three information points among a multitude of such items. They all seem quite credible now, but last summer, they were no more crucial or compelling than dots that indicated bombings, murders, assassination attempts, thefts, cyber-attacks and quasi-military operations. It was from this vast sea that the above pertinent information had to be fished out.

B) The rating of information. Some “dots” are darker than others. The arrest of Moussaui in Minnesota was the darkest dot of the group above - here was an al Qaeda operative who seemed to be plotting something sinister with airplanes. The key here, however, was that he was in custody - and therefore no longer a personal threat. The focus of the FBI was now prosecution of Moussaui, which meant following strict procedures to make sure any evidence discovered could be used in court. Unfortunately, that is often a slow process, and what is found is generally not revealed until the court case proceeds. The vague “al Qaeda hijacking” threat was clearly the lightest dot: Hijacking threats have been almost constant for 20-plus years. The FBI agent’s concern and older foreign intelligence indications fall somewhere in between. A complete review of terrorism assessments over the past five years might very well find there were plenty of “dark” dots which competed with the attention of the “lighter” dots nearby. In fact, the “Millennium Bomber” case, with his information on al Qaeda hijacking and airline bombing plans, uncovered and foiled in 1999, would seem to be a trove of “dark” dots indeed - but they provided no forewarning of Sept. 11.

C) The correlating of information. Ah yes, “connecting the dots.” Now that we see there were a myriad dots of different shades competing for attention, the connections seem much less obvious. That is not to say they could not have been put together; rather, it is an indictment of the system which was - and to a large degree still is - used to find those connections. Think of it this way. You have just been handed line drawings of animals, and told they were the constellations noted by the Inca Civilization in the nighttime sky. Your job is to match them with the stars tonight. Sounds a bit daunting, eh? Obviously, the connections are all right there in front of you; the question is, how do you fit the dots together?

This, I believe, makes clear how nearly impossible it would have been for the government (and even less so a single person) to extract the plot of Sept. 11 from the information that was before it, using the system that was (and is) in place. However, this also points to a new system: centralized information processing and an expert unit dedicated entirely to finding proposed “solutions” to the data points received. Powerful computers could correlate masses of information, then propose possible connections. The experts could look over those connections and rate their probability. Those hypotheses could then be forwarded to an anti-terrorism team comprised of the FBI director, CIA director and director of homeland security. This group would rate the assessments further and provide the information to the president. Obviously, no system will be perfect at rating and connecting the data points, but this centralization, coupled with the creative processes of the expert unit, would provide at least some hope of figuring out the plans of terrorists before they are carried out.

In fact, the whole problem - masses of data which must be understood as coherent groups in order to predict the future - has a lot in common with weather forecasting. Today, meteorologists use powerful computers to correlate huge quantities of information, which then model future changes in the weather. The meteorologists study these “solutions” to the information, then, based both on empirical knowledge and subjective opinion on the patterns involved, they offer the forecasts you and I see everyday. The system employed today is miles ahead of what we had just 20 years ago. General patterns are quite often accurately predicated a week or two (sometimes even more) in advance. Specifics are usually very good a day or two ahead. However, as we all know, even with this system, sometimes the weatherman is wrong. So it is likely to be with terror prediction, where the forces behind all those data points can change their mind and try a different tack at will. Even so, putting our capabilities to full effect on the problem of terrorism could keep us safe from most of the scattered showers of terror - and accurately predict, and prevent, the next big storm.

Comments: Post a Comment