There is a problem with that: It assumes multiple valid responses and deals too much with what 'discover' or 'america' means. It wouldn't work for the 'Why is this steel plate hotter on the side away from the fire' question.
Hermenueutics-like games are risky since I think they teach contrarianness, thinking up unique-non obvious answers w/o regard to correctness. They teach the kind of reason that is rightly accused of being able to argue for atrocities.
I think that's a kind of terrible lesson.
It might work well to come up with a whole bunch of questions that are not to trick-questiony, but in which guessing the password is spectacularly wrong.
For the sciences, a better method might be to set up things where you have to make simple predictions.
As a teacher, I wonder if it is possible to instill this skill into students the skills of rationality and critical thinking. I teach the third grade, and it is not immediately apparent how to apply this with my own class.
The problems I foresee are as follows:
In the sequences, it is suggested teachers should drill into students words don't count, only anticipation-controllers. How practical is this for an elementary school level? Also appreciated would be any ideas or experiences on how to do this, or how to combat the above problems. Hearing from other teachers would be excellent especially.