Still, not jumping in front of the train will 'work' regardless of whether you understand why. And it's a lot more useful than remaining agnostic about train-jumping until you understand the physical principles involved.
Agreed that genuine understanding is more useful, of course -- but correctly choosing which authorities to believe (and with what level of confidence) is often a necessary second-best. How many things do any of us truly understand in full detail?
What if jumping in front of a train ("trainspotting") worked regardless of whether you understood why?
> How many things do any of us truly understand in full detail?
Can you visualize the mechanism? You either do or you don't. Maybe the visualization is misleading. Maybe it's not. But if you're at least doing the motions there, you're revising the mechanism rather than reinventing the wheel from scratch.
> What if jumping in front of a train ("trainspotting") worked regardless of whether you understood why?
Not sure if I understand you here (I don't get the 'trainspotting' reference), but: in a world where not jumping in front of trains was fatal, parents would instruct their kids to jump in front of trains, and the ones who took their advice would survive.
In a world where it didn't matter so much either way, who knows, maybe some parents would needlessly forbid their children from jumping in front of trains. So the kids who didn't blindly follow their parents' advice might benefit from their curiosity. But if they made a habit of ignoring their parents' advice until they fully understood the reasoning behind it, they would do a bunch of other stupidly dangerous things and probably die.
> Can you visualize the mechanism? You either do or you don't. Maybe the visualization is misleading. Maybe it's not. But if you're at least doing the motions there, you're revising the mechanism rather than reinventing the wheel from scratch.
I think there's a big gap between having some kind of mental model, and having a sufficiently detailed, accurate, robust mental model to make independent judgments in important contexts. Sure, a rough high-level understanding can be useful as a preliminary bullshit detector, pinging for things that should be taken with great scepticism pending further investigation; but unless you understand a topic in full detail, you're always at risk of making 'logical' deductions that fail because of unknown (to you) unknowns.
Agreed that genuine understanding is more useful, of course -- but correctly choosing which authorities to believe (and with what level of confidence) is often a necessary second-best. How many things do any of us truly understand in full detail?