My personal belief is that they learn simplified models that aren’t quite correct, without enough explicit warnings that they aren’t fully correct. Then, later on, they’re smart enough to figure out that the models aren’t good enough, so start building their own, without the requisite background of what the physicists’ actual model are, and why certain approaches have and haven’t been taken, and how counter-intuitive experiments have forced certain choices.
It’s like the artillery officer who thinks you can’t use GR to calculate the trajectory of an artillery shell—you have to use Newtonian mechanics to do it. Not that it wouldn’t be practical go use GR, but that it could not be done. He doesn’t realize that not only can you calculate it with GR, but that it would be far more accurate, too (it would, of course, be horribly impractical, however).
My personal belief is that they learn simplified models that aren’t quite correct, without enough explicit warnings that they aren’t fully correct. Then, later on, they’re smart enough to figure out that the models aren’t good enough, so start building their own, without the requisite background of what the physicists’ actual model are, and why certain approaches have and haven’t been taken, and how counter-intuitive experiments have forced certain choices.
It’s like the artillery officer who thinks you can’t use GR to calculate the trajectory of an artillery shell—you have to use Newtonian mechanics to do it. Not that it wouldn’t be practical go use GR, but that it could not be done. He doesn’t realize that not only can you calculate it with GR, but that it would be far more accurate, too (it would, of course, be horribly impractical, however).