As you said that was from preliminary info. In the end the accident wasn't caused by icing. As far as what Colgan pilots were told to do in "icing conditions", we have someone posting on this thread who I'm sure can tell us. We do not disconnect the autopilot every time we turn on the anti-ice systems.
I haven't caught up on the whole thread, but I just wanted to throw in a quick answer to this part of the discussion.
For the record, they both had a fair amount of experience in icing at the time of the accident (not as much as some at the company, but certainly enough to be "comfortable" in it). Their conversation was about how BEFORE coming to the airline, they had very little exposure to icing. That is true of, I'd wager, more than a majority of regional pilots. Most regional pilots these days come from instructing. Most instructors don't spend a whole hell of a lot of time slogging around in light to moderate icing all winter. Certainly not as much as regional turbo-prop pilots do.
Before the accident, CJC had no stance on AP usage and icing. Bombardier had a limitation in the AFM (and in our CFM) for no AP in SEVERE icing.
After the accident, CJC training said, essentially "you get a better feel for the plane without AP on." Well duh. But still the only limitation was for severe icing.
I flew the Q for a little over two years, including on the night of the accident, and I never did myself, nor witnessed anyone else, turn off the AP just because we were in icing. That plane was a PITA to hand fly for long periods of time, especially up high; the AP was your friend in that thing.
Icing had nothing to do with the accident. It may have been a brief distraction, but really their conversation about icing holds no more significance than a conversation about ketchup on hot dogs. That plane will carry a LOT of ice...
The tail stall conundrum is a result of CJC training at the time. All initial training and annual recurrents included the FAA's rockin' video from the '80s about tail icing/tail stalls. FAA said we had to watch it, so we watched it; every year. Most of us had that stupid thing memorized. Marv's reaction fit almost perfectly with the response dictated for tail icing/tail stall (minus the whole, not saying a thing).
We didn't find out (from Bombardier) until months after the accident that the Q cannot tail stall. At all. And with a purely hydraulic elevator, elevator "snatch" (what could be confused with the pusher) is physically impossible.
Jesse - I grant the premise that Marv should have said something if he was doing a non-standard (non-trained) tail stall recovery procedure. BUT, the normal stall recovery procedure is supposed to include a verbal, Pavlovian response to the 6 indications (is it 7?) in the cockpit that the plane is in a stalled condition. The fact that the stall protection system worked as advertised, but Marv (and Bekki, for that matter) were almost completely silent, indicates to me that either 1) he thought tail stall and simply had no callouts for it, 2) he forgot all the callouts for something we practice ad nauseum every 6 months, or 3) they had lost all SA and had no idea WTF the plane was doing. I suspect the truth lies somewhere between those options, and perhaps several others that we will never fully understand.
There were a lot of other issues at play that night WRT the airplane's icing systems, system integration, training, manual content, procedures... Human error was the final straw, but as with all accident, there was a long chain of events that led up to that house that night. As far as any of us could tell after the fact, icing wasn't on that chain.
