Left I on the News

The headlines cry out:

Panel warns biological attack likely by 2013

"Likely"? That's not a scientific term, but according to another report, the opening sentence of the summary says such an attack "is more likely than not," which says there's at least a 51% chance of such an attack.

And on what is this speculation based? Well, the report is classified, but based on the news reports, the usual precious little. We're told "the threat of a terrorist attack using nuclear or biological weapons is growing "not because we have not done positive things but because adversaries are moving at an even faster pace to increase their access" to those materials." They are? Considering the evidence (none) that even the alleged "biggest threat" of such an attack (Iran) is developing nuclear weapons, I'd judge it likely that the evidence that Al Qaeda or anyone else is "moving at an even faster pace" to increase its access to nuclear or biological weapons is again precisely none.

How reliable is that ">50% likelihood" number? Consider this sentence:

The report of the Commission on the Prevention of WMD Proliferation and Terrorism, led by former Sens. Bob Graham of Florida and Jim Talent of Missouri, acknowledges that terrorist groups still lack the needed scientific and technical ability to make weapons out of pathogens or nuclear bombs. But it warns that gap can be easily overcome, if terrorists find scientists willing to share or sell their know-how.

So, all we have to do is to estimate the probability of "finding a scientist willing to share or sell their know-how." I'd like to find a scientist or mathematician willing to share their know-how on how you estimate the probability of "finding a scientist willing to share or sell their know-how." Obviously you can't, it's pure speculation.

Likewise, there's this:

Al-Qaida remains the only terrorist group judged to be actively intent on conducting a nuclear attack against the United States, the report notes. It is not yet capable of building such a weapon and has yet to obtain one. But that could change if a nuclear weapons engineer or scientist were recruited to al-Qaida's cause, the report warns.

How they "judge" that Al Qaida is "actively intent" on such an attack I have no idea. But again, the idea that even if that were their intent, and even if given a weapon they could set it off in an appropriate location (not a trivial matter), we still are left with the impossibility of estimating how "likely" (or unlikely) it is they could recruit a nuclear weapons engineer. And is one engineer or scientist really enough? Countries that develop nuclear weapons have hundreds, probably thousands, of scientists and engineers working on the problem, in real laboratories filled with scientific equipment.

I'm not saying countries aren't wise to take measures to secure nuclear and biological materials against theft and subsequent misuse. But this article, and it's claim that such an attack is "likely" by "2013" (really? Is it only 40% probable by 2012? What's the probability by 2014? 2015?) is an abuse of the scientific method, pure scare story, and pure balderdash.

Of course, no one should forget that the biggest bioweapons attack in recent years, the anthrax attacks, came courtesy of the U.S. government (whether we know the actual perpetrator or not).