"Nobody I knew, including some avid Airbus Design-Flaw Seekers, ever doubted that the dynamics of the aircraft as shown on the FDR were highly inappropriate. The debate centered on whether one could conclude that the pilot did it (rudder pedal travel was not recorded on the FDR) or whether some automatic system had gone awry. Obviously, the NTSB's attempts to find some rudder control system that had a failure mode that caused such motion came up with nothing. (I am inclined to think that if the NTSB cannot find one, then nobody can. They have a lot of experience and succeeded in finding the failure in the Parker-Hannefin yaw damper in the B737 after the US B737 crash in 1994...
The NTSB is finding that indeed the pilot did it. All else follows from that."
It's folly to assume the NTSB is immune from political pressure. The USAIR B737-300 crash in 1994 followed a UAL 737-200 crash in Colorado Springs in about 1989, which made no sense until the rudder control reversal was admitted to in 1994. That was in spite of a UAL 737-300 going into rudder control reversal on taxi out in about June of 1991. After that was the PBS special on "The Mysterious Crash of Flight 301", a COPA 737-200 that crashed in the Panama jungle one night in about 1992. Some insiders blamed the Sperry vertical gyro, but to me it had all the earmarks of the Colo Springs crash: rudder control reversal. Five years and three accidents does not inspire confidence the NTSB got it right on AA 587.
GB