Also what AA found out in "discovery" (legal teams) is that the particular accident aircraft A300 had been involved in an inflight upset some time before, and that the flight data recordings from that upset were given to Airbus. In internal memos between Airbus engineers who communicated internally with themselves that OMG, that very A300 vertical stabilizer had experienced ultimate design load factor in it's inflight upset. That revelation was never passed on to AMR. There was also damage to the empennage attach fittings on the A300 that occured before the tail was attached to the plane.
Also, 411 you may want to check the NTSB recommendations (never complied with BTW) of A300 oversensitive rudderpedal design "logic".
A day or two later, when walking in the debris field at the crash site, our chief accident investigator was along side his opposite from Airbus when the two of them came upon the smoldering vertical stabilizer attachment fitting from which the tail ripped. Our investigator asked the Airbus guy. "In flight test, do you guys perform rudder doublets?" The incredulous response was, "No, the tail would come off." The test is full deflection, right or left then relaxed to center. Brief pause, then rudder deflection the other way.
I don't think there were many transport jet pilots around prior to 2002 who gave much thought to how easy it would be to take a rudder, no make that the ENTIRE FIN off a modern jet, well, at least an Airbus fin with a few rudder control inputs and at relatively low speed. I know, I never gave a thought about it on other than smooth use of the rudder in the Boeings and Mickey D's I flew. After all, we were told all about rudder limiters, yaw dampers; and the new fangled Airbi, who would take care of dumb pilots and not allow any over the limit control inputs. Not funny, but I've seen lots of detached Airbus tails floating in water, just one of them from an AA aircraft.