TC
perhaps we need to consider what is the model for success (of anything), before we consider a model for failure.
Would a model for success have a series of elements, all interacting in a linear sort of way, that lead to a successful conclusion ?
Would a failure model be the same, but with one component that is changed/modified for whatever reason, in whatever way,which then gives a different, less than expectation result.
I'd keep a wide open mind on this, and try not to be to aviation specific. Good question for us all to ponder. Worth starting a seperate thread on this ?
BS