Research highlights that startle and surprise are similar, but surprise is more significant due to duration and depth of effect. Also, that surprise can build slowly or be reinvigorated if events do not progress as expected after crew action. (Landman).
The description of cognitive effects is a good guide, but the conclusions of the research into methods of intervention are less convincing; similar to the range of views in this thread.
Opposed to debating alternatives to alleviate the issue, reconsider what the issue actually is, why is surprise now being considered - a new fad.
Surprise is a human reaction to a situation - stress, part of our natural defences and education.
Why is surprise now of greater concern in aviation; human nature hasn't changed.
Aviation is evolving, safer, fewer challenging situations from technical failure; similar due to the environment. Thus the view that decreased exposure to surprise and experience of managing the effect is contributive.
Conversely, historical accident rates were higher than now, but we did not record the number of accidents involving surprise. The contribution of surprise then could have been very much higher than now, so in effect aircraft and environmental change has helped - even thought humans will always be surprised. The issue is that we are ‘surprised’ by the apparent change or not knowing why.
Another view is that with high levels of safety, accident investigators have fewer occasions to probe human contributions, but when opportunity arises they do so in more depth. Very few accident reports cite surprise as ‘the cause’, only as a possible contribution, which investigations consider with different levels of debate along with other possibilities.
The issue is typically complex, many views of what should be considered or implemented. Also those who profess solutions may only contribute to the mess; regulators, training organisations, etc.
Re-reading the research on training intervention with a view to identify the issue opposed to intervention, suggests that much of what is already being trained could be detrimental.
Regulators could review what is being trained now, and how this might relate to research findings - negative training; also to consider that by requiring more training they risk diluting hard pressed resources from more urgent safety need.
Landman
https://pure.tudelft.nl/portal/files...on_startle.pdf