This Probably Has Merit, Possibly
Interesting piece at The Chronicle of Higher Education. I distrust any analysis that divides people into two classes, but such an approach is often useful to understand a new way of understanding something.
Clarke divides people into probabilists and possibilists. Much modern scientific and governmental policy about disasters, he claims, emerges from probabilistic thinking -- "What's the likelihood that the nuclear plant will melt down?" -- while possibilistic, or worst-case, thinking asks "What happens if the nuclear plant has a really bad day?"
Clarke asserts that we engage in worst-case thinking as individuals every day -- we buy insurance, decline to take up sky diving, and so on. But when risk assessment broadens from individual decision making to societal setting of policy by "elites and institutions," probabilists rule, and too often stigmatize possibilists as irrational. . .
The important truth, Clarke insists, is that chance "is often against us." As a colleague he quotes says, "things that have never happened before happen all the time." Society needs to be protected from the possible, not just the probable. Too often, Clarke argues, "probabilism" protects the powerful to such an extent that "reasonable" simply "means probability."
"I am not an alarmist," Clarke confides, "but I am alarmed."
Clarke is not, truth be told, wholly fair to probability theorists. The best recognize that there's no valid inference from probability judgments to judgments of acceptable risk, that the latter imply moral and evaluative judgments. . .
Worst Cases analyzes many notorious disasters, among them the Black Death of medieval Europe, the sinking of the Lusitania, the destruction of Galveston by hurricane in 1900, the Challenger shuttle disaster, the Mississippi River flood of 1993. Clarke details scenarios waiting to happen, such as terrorist-sponsored outbreaks of Marburg or Ebola virus, or the effect of a Category 5 storm on the two nuclear plants 25 miles south of Miami. In all his hypotheticals, purely probabilistic thinking augurs hard times ahead. . .
Probability theory and risk analysis may not be the jazziest topics to wrestle with in Katrina's wake, but more should try. And not just in regard to the Gulf Coast. Maybe it is too soon to worry about Asteroid 1950 DA, which Clarke advises has a "credible chance" of striking the earth on March 16, 2880. But what about the 180,000 pounds of sulfur dioxide stored near Kearney N.J.? And the potential earthquake under Manhattan? (Yes, there's a scientific group that studies it.)
Link.
