Dead stick and fixed throttle.
If you "suddenly wake up" after a jolt or bang, then the first pilots instinctive action is to grab the controls. In a Boeing you will get instant tactile feedback that "it" turns to ... and that the thrust is up high or low. Every research found out that the tactile input is like parallel computer input, 10 times faster than a serial input. Humans react similarily. If you have to collect data via intellectual input (visual on display) you finally get the same values, agreed, but 10 times slower. That's my point.
I would go as far as pretending that Habsheim and Bangalore would have been avoided in a Boeing, because the tactile feedback, however distracted the pilots may have been, would have instinctively told them "...there's NO thrust" and they would have slammed the throttles to the mechanical stop. Any excuse like "they fiddled with the cb's or with wrong modes" is oblivious: If a pilot induced accident could have been avoided by better design, there's no excuse NOT to implement it.
Furthermore there is the pending outcome of the investigation of the Toronto accident. Apparently the inspectors will argue that a disconnection of the AT would have been the better solution. If however you disconnect the AT in a Airbus, you lose some protections. In contrast on a Boeing under such circumstances you need not disconnect the AT. Even if it works badly, by overcorrecting huge gusts, you can leave it in as a valuable back up and protection, but you can hold it back or override it with thrust, if required. This is undoubtedly the much better suited solution and again, there's no excuse NOT to implement a better one for extreme conditions. Especially if the constructor and the operators mostly recommend to use the AT.
That's my point.