Comment by PaulHoule

One of the major points in that book (missed by a lot of people who read it seemingly and certainly missed if you read about that book) was that NASA had a system for normalizing deviance officially in that they knew the Shuttle was unsafe at any speed and had numerous unacceptable safety flaws. For each launch they had a meeting where they reviewed all the things that they knew were wrong and made the decision that the risk was acceptable or could be managed and would go ahead anyway.

When the O-ring problem was presented at this meeting it was just another in a long list of things that could blow up the shuttle and wasn't all that remarkable among them except that, in retrospect, it really did blow up the shuttle.

That is, NASA's normalization of deviance wasn't like the individual worker who comes to the conclusion that it is OK to smoke pot and weld, but rather was an official process in the organization. A similar thing is seen in other safety critical fields. If you build a nuclear power plant with two units one might be the mirror of the other except you forgot to mirror how the door to the employee break room opens so it opens the wrong way. Anything complex like that it going to wind up with numerous deviances of that general nature and it is the job of a committee to file out paperwork on each and every one of them to decide what is going to be done about it from nothing to reworking it or taking some remedial measures in operations.