framer;
I apologize to you and to KAG for my complete lack of clarity in my response. I failed to make my meaning of the notion, "knowing", and therefore my point, clear, because I wrote the response in a hurry - always a mistake! Permit me a second shot at it:
The comments were in response to your post which I thought was very interesting and posed and interesting question. Nothing "directed at" you, just a response about CRM-type situations and how the medical profession could make progress in adopting, with substantial modification (for reasons given, ie, the surgeon is not the "captain", etc) such human-factors processes.
KAG, you're correct in saying "safety is the state of being safe", and that "knowing a wing is on fire" doesn't make things "safer". I can see where the confusion arose!
The key point I was attempting to make was, flight safety is about "what" not "who". It was trying to acknowledge that "knowing" (which is about 'what'), promoted understanding and therefore a safer operation, than "who", which is about someone's opinion, usually subtlely grounded in "authority" or "position", of how things should be run and which rarely produces understanding because it seeks obedience, not questions.
The difference is subtle. Obviously, knowledge and opinion may be the same in many circumstances but one is grounded in information available to the crew which someone is referencing, and the other appeals to a position of authority.
CRM is about information exchange, or "knowing", without regard to who was conveying the information. Dismissing information based upon who is conveying it is the exact opposite of what is intended by CRM.
Let me describe it in very black-and-white, simplistic terms, realizing that most issues and circumstances are far more subtle than this.
Previously, (pre-CRM type communications in cockpits), if a junior, (read "inexperienced") member of the crew said something, the crusty old captain might dismiss it and even think that the guy in the right seat was trying to take over his airplane.
F/O: "Captain, should we be descending to the procedure altitude out here?"
Captain: "He cleared us for the approach, didn't he?!"
F/O: "Yes, but we're thirty miles out..."
Captain: "We're cleared for the approach and that means we can descend!"
F/O: "But the chart... ,"
Captain: "I've been in and out of here hundreds of times. It's fine..."
In this admittedly-extreme example of a conversation, the decision-making process is about "who", (the captain vs the F/O), not "what", (should we be descending...?" The captain dismisses the F/O's comment subtlely using his position and claiming his experience with the airport which is about "who", to justify the descent instead of going further with the F/O's intervention and determing "what".
These circumstances applied at Tenerife, at Buga, (Colombia), Little Rock (Arkansas) and dozens of other accidents. CFIT accidents typically though not always, begin with poor CRM. The F/O on the KLM B747 questioned the captain's decision but he elected to take off, dismissing the intervention.
What I meant by my comment about "junior people flying the airplane" is, it is a serious misunderstanding of CRM on the part of both crew members to think that CRM is equivalent to "the F/O is taking over the airplane" if he or she raises questions about the operation. Questioning the captain is not the same as trying to fly his/her airplane. The captain is always in command by law, (one significant difference between aviation and medicine), and will always make the final decision unless it clearly threatens the safety of the flight. That does not relieve the other crew members of the duty and responsibility to speak up in all circumstances that require "knowing" about a situation.
Far more often however, it is not a "beligerant" captain but a captain who may have made a small error, didn't know something or assumed something and other crew members were either afraid to speak up out of embarrassment of possibly being wrong, or out of respect because the commander was "such a good pilot" or possibly out of fear because the commander was known to "bite off heads" if challenged. The dynamics are as varied as people are but the common thread is, "nobody said anything" and so there were two or three different understandings of the operational situation riding in the cockpit instead of one.
CRM came about because accident investigators realized through the CVR that a perfectly serviceable airplane was flown into the ground or into another aircraft killing a planeload of people because communications did not effectively establish a full, common understanding of the situation.
CRM doesn't end just with understanding.
One example of CRM in action was the United Airlines DC10 at Sioux City, where Denny Fitch, a dead-heading pilot sat behind the captain and First Officer and manipulated the throttles to fly the aircraft to the Sioux City airport. The captain handed control over to someone not even in his crew but who was "best qualified in the moment" to do the job; - "What", not "who". It was a brilliant command decision that saved 185 lives.
I hope this is a bit better explanation of what I was trying to say and that the simplifications are not otherwise patronizing your aviation knowledge. The subtleties of CRM and what CRM has evolved into now would take a much longer post but I hope the idea comes across reasonably well. I think medicine could benefit tremendously from such an approach, though, (as stated), modifications to suit the unique dynamics of medicine (vs aviation) would have to be made.