|
[an error occurred while processing this directive][an error occurred while processing this directive][an error occurred while processing this directive][an error occurred while processing this directive] |
  
|
|
Columbia: a `normal accident'?
By Alex Beam, Globe Columnist, 2/4/2003
Human guilt, or at least responsibility, is a concept that is easy for us to understand; it is the underpinning of the Christian religion and of a lot of Western philosophy. But in thinking about the shuttle accident, the one factor that can be eliminated almost immediately is pilot error. The craft seems to have been on a sophisticated version of autopilot when it disintegrated. The shuttle catastrophe looks a lot like a ''normal accident,'' a term coined by Yale sociologist Charles Perrow almost 20 years ago in his book ''Normal Accidents: Living With High-Risk Technologies.'' Perrow's point is that large, complex systems can fail by themselves, without human intervention. ''If the system is large, the possible combinations of failures are practically infinite,'' the pilot and journalist William Langewiesche wrote in the Atlantic, applying Perrow's theory to a commercial airline crash. ''Such unravelings seem to have an intelligence of their own: they expose hidden connections, neutralize redundancies, bypass `firewalls,' and exploit chance circumstances that no engineer could have planned for.''
Perrow himself isn't so sure that Saturday's events fall neatly into his well-known template. If, as he suspects, the shuttle's insulating heat tiles failed on the left wing, ''it's odd and almost embarrassing, but this may be a case of a single point failure, not a complex interaction of multiple failures,'' he said by telephone from Frankfurt. Perrow is quite critical of NASA's overall commitment to safety. ''We are bound to have occasional accidents with such a complex system, but we've been running the risks higher and higher. The stuff that's coming out now makes it sound like a replay of the Challenger explosion: They were cutting back on safety, claiming they had budget problems while the General Accounting Office was saying that they were mismanaging the contractors.'' NASA has never invited Perrow to speak or consult at the agency, although its engineers seem to be well acquainted with his work. You can read an intriguing NASA presentation on ''normal accident'' theory on the Web at www.hq .nasa.gov/office/codeq/accident/ ''I was horrified to learn that their top administrators had reinstituted the teachers-in-space program, because it means the people at the top had lost sight of the risk of operating an experimental program,'' she says. After the Challenger crash, she notes, it emerged that lower-level NASA staffers were more cognizant of potential danger than their bosses; ''they were witnessing all kinds of things going wrong.'' Aside from their shared interest in the failure of complex systems, Vaughan and Perrow have something else in common: the cold shoulder from NASA. ''When my book came out, everybody with any interest in risk called me. The US Forest Service called me -- even my high school boyfriend called,'' Vaughan jokes. ''But NASA never called.'' Alex Beam is a Globe columnist. His e-dress is beam@globe.com.
This story ran on page E1 of the Boston Globe on 2/4/2003.
|
|
|
|
|
|
© Copyright 2003 New York Times Company |
|||||||
istorically, when an airplane crashes, there has been tremendous pressure to blame ''pilot error.'' Pilots resent such conclusive leaps, and their union monitors accident investigations closely. But blaming the human at the controls furnishes the shortest, most comprehensible explanation for extraordinary events -- like the Columbia disaster -- that can be impossible to fathom.
