Will;
I think too, that a loss of sensors as is demonstrable prior to if not a part of this accident, presents an arguably suitable example of the differences involved in manual flight and a litany of pre-programmed failures that cascade from a manageable defect into a very dangerous reluctance of the computers to perform at even a very basic level of expectation.
First, I am not discussing the AF crew here or suggesting anything in the following points.
While I remain cognizant of the B2 accident at Guam, loss of sensors/sensor data does not, in and of itself, generally result in loss-of-control resulting in an accident. The QRH Unreliable Airspeed/ADR Failure drills are in place for such a loss and have been used to successfully maintain a safe flight path while recovering the necessary sensors. I have had this experience in a 767; there are recent incidents which attest to this understanding. Given only what we have, we can assume that the flight control computers did not stop functioning and that they were still controllable by the stick even as they reverted to alternate law even with the loss of PRIM1.
There is no special technique or training one needs for flight in alternate law; training
is required for flight with unreliable speed, however, but such training, with shortened simulator times, (down to 3.5hr sessions in some cases) is rare.
In the absence of ACARS indications, we could conclude that it is not these computers which were in the cascade of failures; from what little is known from the ACARS, we can safely assume that the flight control primary and seconary computers (FCPC/FCSC), minus PRIM1, continued to function as expected at least until 02:14:59Z.
In 99.9% of all mode confusion occurrences on this and other automated aircraft, "What's it doing now?" is a statement about training and knowledge, not about the aircraft itself. Millions of hours of safe flight and millions of successful departures and arrivals attest to this. The 320/330/340 series aircraft can be hand-flown just like a 172. Not understanding what your aircraft 'is doing now' means an instant disconnection of the autoflight system until one has regained situational awareness inside and outside.
I think we probably agree if I am reading your statements as intended, that this isn't primarily an "airplane" problem. I think there are important things to say about 'pre-programmed' cascading failures but redundancy has it's own issues (as per a recent (linked) paper on the subject), as does a system which contemplates monitoring/judging like a human being to self-assess, "What am 'I' doing now?"
The apparent failure of a pitot-static system and the apparent consequent loss of possibly 3 ADIRUs (with attendant faults/failures) do not render the flight controls inoperable/uncontrollable - that's my only point. Again, I make no judgement here - I am merely unpacking the statement a bit.
Instead of a 'tool' the automatics are proposed to be nothing less than a superior 'pilot', whose purpose is to 'protect' the airplane. Inherent in this language is an 'attitude' (a human one) of 'supremacy'.
Well said. That was my initial impression when I first began to fly the 320 in 1992 and, while one learns one's aircraft as thoroughly as one can and I found the Airbus series a beautiful aircraft to fly, that impression, bolstered by some experiences, never fully left me. It is why I both defend the airplane and criticize it.