November 8, 2002
The discussion about the AA crash over Queens seems to be giving rise to a debate that has more to do with geopolitics than with safety.
Airlines are not generally known for wanting to invest in instrumentation (BA is the exception that springs to mind). We can discuss scenarios where pilots applied excessive rudder input until we're blue in the face: Unless the FDR shows it we just won't know.
Manufacturers can get away with their quaint theories on pilot error because the airlines don't demand FDR improvements that would just tell what happened with the controls and control surfaces in sufficient detail.
Several comments I've seen on AVweb and on other forums seem to focus on the fact that the tail was ripped off this airframe. Frankly, I don't think it is worse that an airliner comes down because rudder failure causes the tail to be ripped off, than that an airliner comes down because of rudder failure with its tail intact. What I do care about is that, from the evidence the public has seen to date, rudder related crashes cannot adequately be explained from the FDR data, simply because not enough data is recorded. And worse, because historical data on that aircraft isn't kept.
I'm an instrumentation buff, I'm the first to admit that. But without improvements in that area I'm afraid we'll be playing the blame game for a long time to come, and I feel safe to predict it isn't going to be the manufacturers or the airlines that will wind up holding the hot potato.