my model suggests that everyone's fixating on language X's expressiveness is a red herring - the real issue is its lack of error handling, which is a much more insidious problem than just "not being as expressive" as someone else