bear;
This is a philosophy of controlled flight I still don't ken
Frankly, when I first stepped into the A320 cockpit in 1992 I felt as though a "muslin veil" was between me and my airplane. I was experienced at glass and FMS use from the B767 and L1011, (brilliant airplane, perfect blend of automation and bread-and-butter solutions) but
this was really different. The training was pretty tattered around the edges - good but not deeply knowledgable. When I got to fly the airplane, it was a non-event. I really liked the solutions to the problems of daily operation, (ILS intercepts, too high on the approach, flight-path-angle solutions, warnings and especially the ECAM which I thought was exceptionally well-considered but not perfect). I always thought there was too much programming of the FMS for good SA and that there was far too much talking and "process" when dealing with a complicated failure, (dual hydraulic, emergency electrical config, etc) and I know very well that with a rapidly degrading Airbus, the circumstances can be extremely demanding, perhaps too much so. Self-diagnosis was still in its infancy - perhaps the B787 will be better, leaving the crew to "bread and butter flight", just as it should be.
The key, at least for me, was to disconnect the airplane when up against one of those (at the time, frequent) moments when we didn't know what the autoflight was doing. We didn't call her Capricious Christine for nothing. Handflying the airplane including manual thrust was like flying the DC9 or any other B&B airplane...it was a non-event and entirely transparent to the task at hand.
The area you may be struggling with may be those black swan areas of software and airplane design where airline pilots almost never-in-a-career may find themselves. The boundaries of controlled flight are not areas where airline pilots ever expect to be; since the jet-age began in the late '50's, it has been a By-the-book industry and those who go beyond for whatever reasons and sometimes not through their own hand, pay the price whether it is turning off the yaw damper at high altitude, (a form of automated intervention which no one seems to claim "interferes with the pilot's control of his/her airplane) and watching dutch roll unfold, to climbing their aircraft to its absolute service ceiling and watching as the speed bleeds back and the engines flame out and suffer core-lock, to ignoring perfectly serviceable cockpit warnings from tested systems such as EGPWS and TCAS.
Put succinctly, the boundaries of software-initiated interventions do not limit the pilot from doing what she or he can do with his/her airplane until the very limits of controlled flight are reached. Why then, is that so unusual? Why is it even controversial? In philosophy, the argument resembles the "how many angels on the head of a pin?" argument. I say this not in debate but in recognition of the millions of hours of successful, unremarkable flight that all designs have amassed without incident or fatal accident.
I have experienced enough failures, one or two quite serious, in operation in the Airbus product to know it performs very well providing one knows one's airplane, just like any other aircraft. Of course, my experience is only that and cannot claim by itself to reify a trend, but it does somewhat mirror industry experience.
I believe that skills and familiarity with autoflight's operation and handling, in and of itself, can (and should) be placed within a different cognitive category than the cognitive skills required in manually flying an aircraft. This isn't the place to pursue that discussion, but I believe this is not a technical matter but a psychological one. I think we are examining the wrong areas using the wrong tools in search of a solution to "automation's interventions" and the sense that we are competing with the computers when flying our airplanes. I have been wondering if the Islamabad accident may be the very case in point which may best illustrate this, but we will have to wait for the data before theorizing further.
If however, as the article states as it 'quotes' those who have listened to the CVR, that heading was used to turn left and nothing further was done in terms of disconnecting everybody and flying the airplane in an escape maneuver, then they may have been stuck in frame of mind where "automation solves all problems" and not "psychologically" sitting in an airplane with a mountain ahead of them. If so, I think this may be a key to delineating some solutions to the human factors issues which are often expressed regarding automation's veil.
Just wondering...
PJ2