By one possible definition, this is the difference between science and everything else like medicine or engineering: science prioritizes advancing knowledge above all else, so (in a system not distorted by publication metrics) that means it prioritizes transmitting and in particular saving it.
Engineering does not—because of commerce, engineering can often even be actively hostile to transmitting knowledge. (I have a person next to me who knows how enterprise SSDs work and refuses to tell me more than “with great difficulty and ingenuity”.) Even without that, the task is difficult enough that without an explicit pressure to refine and preserve explanations that just won’t happen. There always will be people who value and enjoy that, but without a system in place to cultivate and reward them, there won’t arise that a kind of “collective knowledge” akin to collective immunity (or percolation) that the gaps can filled in.
This is actually a problem in the more engineering-oriented parts of science as well. For example, this is one of the problems with Hossenfelder’s suggestion that we stop building colliders: if we do that, in ten to fifteen years (a couple of generations of grad students) it’s fairly certain that we won’t be able to build them on anything like the current level. The US has already experienced this kind of institutional knowledge loss, in fact, with the shuttering of the SSC. So while I agree that they haven’t produced new—not just knowledge, but understanding in quite a while, we need to be really really sure before agreeing to lose this expertise.
And yet, look at physics. Like take a quantum field theory course (taken by what, probably a thousand, ten thousand students per year?) and trace the dependency chain. Even discounting all the backround knowledge we take for granted (e.g. “electricity exists”), the dependency chain is really really deep, probably something like a decade if you count the requisite high-school parts. (It includes some things that might seem unrelated on first glance but are in fact essential, like optics and thermodynamics.) That creates an actual problem in teaching it. And the subject itself is a year to two years deep at least before it becomes wide enough that you can mix and match topics, to say nothing of actual research.
I shudder to think what would happen if you tried to work through all of the materials science you encounter in an average first-world home (and all of the materials science, metallurgy, chemistry required to build that, etc.). An extremely fun project to contemplate, but I suspect too long for a human lifetime.
Engineering does not—because of commerce, engineering can often even be actively hostile to transmitting knowledge. (I have a person next to me who knows how enterprise SSDs work and refuses to tell me more than “with great difficulty and ingenuity”.) Even without that, the task is difficult enough that without an explicit pressure to refine and preserve explanations that just won’t happen. There always will be people who value and enjoy that, but without a system in place to cultivate and reward them, there won’t arise that a kind of “collective knowledge” akin to collective immunity (or percolation) that the gaps can filled in.
This is actually a problem in the more engineering-oriented parts of science as well. For example, this is one of the problems with Hossenfelder’s suggestion that we stop building colliders: if we do that, in ten to fifteen years (a couple of generations of grad students) it’s fairly certain that we won’t be able to build them on anything like the current level. The US has already experienced this kind of institutional knowledge loss, in fact, with the shuttering of the SSC. So while I agree that they haven’t produced new—not just knowledge, but understanding in quite a while, we need to be really really sure before agreeing to lose this expertise.
And yet, look at physics. Like take a quantum field theory course (taken by what, probably a thousand, ten thousand students per year?) and trace the dependency chain. Even discounting all the backround knowledge we take for granted (e.g. “electricity exists”), the dependency chain is really really deep, probably something like a decade if you count the requisite high-school parts. (It includes some things that might seem unrelated on first glance but are in fact essential, like optics and thermodynamics.) That creates an actual problem in teaching it. And the subject itself is a year to two years deep at least before it becomes wide enough that you can mix and match topics, to say nothing of actual research.
I shudder to think what would happen if you tried to work through all of the materials science you encounter in an average first-world home (and all of the materials science, metallurgy, chemistry required to build that, etc.). An extremely fun project to contemplate, but I suspect too long for a human lifetime.