I've pasted in an excerpt from a paper I wrote while working on my Master's...thought you might be interested.
As a flier, I have issues with Airbus due to the responsibility-authority double bind discussed below; as a passenger, both Boeing and Airbus have compariable accident rates so I'm happy to fly aboard either.
As summarized by Wallace (2000), the differences between the two competing manufacturers’ automation can be defined by who has final authority over the flight controls when the aircraft is near the edge of the flight envelope—the pilot, or the computer. Airbus uses hard limitations to determine into what flight regime it will allow the pilot to command a plane, while Boeing utilizes soft limits, allowing the pilot to override computer limits (Wallace, 2000).
Hard limits, as utilized by Airbus, mean that no matter how much control stick input is made, the flight computers will not allow the aircraft to exceed any of its predetermined boundaries for pitch, roll, g-force, and the like (Ropelewski, 1996). Hard limits can prevent an aircraft from executing maneuvers at its true maximum limitations, as structurally, aircraft tend to have higher limits than noted in their documentation (Wallace, 2000). Boeing uses soft limits, which have many of the same limitations on flight regime, but allow the pilot to override the limitations with extra force on the yoke or throttles.
Olson (2001), writes of the responsibility-authority double bind inherent in hard limits—the fact that a pilot can be held responsible for the safe operation of his or her aircraft without having full authority over the controls, such as under Airbus’s design philosophy. He indicates a necessity to anticipate conflict, as the pilot may not have the authority necessary to intervene after an error has occurred. This, and an inability to design out all instances in which a pilot would need to override the computer, coupled with the inability of the pilot to anticipate all instances of conflict with the FMS, gives rise to increased potential for incident.
In a differing opinion, Last (1989) describes the flight envelope, or ‘hard’ protection feature of the A320 as the aircraft’s greatest benefit to its pilot. In his words, “initial reservations about manoeuvre limitation in extreme situations seem on balance to be more than countered by the freedom to apply maximum input without concern” (p. 1486).
Bent (1997) suggests that it is often business practices, not necessarily safety, which drive the manufacturers philosophies toward automation. He recommends commonality in certain areas of automation design to benefit safety, but suggests there is no overall right way to automate aircraft; he states aircraft procurement by air lines is driven by airframe efficiency rather than automation architecture, and therefore commonality between aircraft manufacturers could increase safety without affecting profits. He suggests manufactures intentionally design automation schema with the sole intent of being different rather than better, to the detriment of commonality and safety. According to Wallace (2000), the pilots union has issued recommendations that the best of Airbus and Boeing designs be integrated into future designs.
Bent (1997) states public perception of highly automated aircraft contains many unfounded fears with regard to passenger safety; historically neither has the advantage. Confrontation between the differing design philosophies, according to Bent (1997), can usually be traced to competition between manufacturers. Statistically, Boeing and Airbus maintain accident rates that are equivalent (Wallace, 2000).