In proper physics, by contrast, you need to write down an equation that applies in many different situations and stick to it. It’s gotta have variables with specific definitions, it’s gotta have a specific domain of applicability, etc. Everything has to be specific, specific, specific—so specific that in any conceivable situation, there is a right and wrong answer to the questions: “Does the equation apply here? And if so, what exactly do the variables mean in this context?” That’s how you know that you’re not making things up as you go along.
Anyway, Newton’s laws are in fact proper physics, hence there’s a right and a wrong answer to how to apply Newton’s laws in any given situation. Sometimes people do it wrong, and sometimes they only notice that fact after running the experiment. But when that happens, then they should be able to find a locally-invalid step involved in the original prediction.
The claim “There’s a right answer to the question of how to apply Newton’s laws in any given situation” is a claim that Newton’s laws are meaningful, not a claim that they’re true in the real world. (Cf. “correctly” applying Newton’s laws to predict Mercury perihelion precession.) As such, you can verify this first claim without even running any experiments. E.g. get a bunch of good physicists, and show them all a bunch of local derivation steps in Newtonian physics problems, where some of those steps are locally-valid and others are locally-invalid, and see if the experts can figure out which is which. (Maybe you could make it easier on the experts by providing an argument for each side, a la “debate”.)
If Newton’s laws are meaningful (and they are), then it is at least possible in principle to compare their predictions to experiments. Then there’s still a separate question of whether the community of scientists and engineers are in fact capable of reliably doing that.
Luckily, every day around the world, there are thousands of new examples of honest-to-goodness predictions using Newton’s laws that turned out to be correct on the first try. That observation conveniently validates both the hypothesis that Newton’s laws are meaningful at all, and the hypothesis that scientists and engineers are collectively up to the task of applying Newton’s laws correctly, at least when they really set their minds to it.
At my old job we would sometimes be in a situation where we really needed our physics calculation to agree with experiments the first time. It was always pretty harrowing. Lots of people checking each other’s work and so on. Even so, the customers were always very skeptical of such claims—and probably justifiably so.
I worked for five years at Draper, an R&D lab mostly in the USA military-industrial complex. I did a lot of calculations of the form “if we build this kind of gadget (usually a sensor), will it to actually meet the performance requirement?”. I was usually involved at a very early stage, when there was a wide space of possible design decisions and we hadn’t committed to anything yet, let alone started prototyping etc. As the project proceeds, you ramp up to more and more detailed models that make fewer and fewer assumptions, and you start supplementing that with prototype data and so on… but meanwhile the project costs are growing exponentially and it becomes almost impossible to make any more big design changes. So the calculations needed to be as faithful as possible, right from the earliest BOTEC / spreadsheet stage. No factor-of-2π errors allowed!! I think I was good at it … or at least, if I screwed anything up, nobody ever told me. :)
The firm also had projects that for things like designing & building actual gadgets that would then actually get launched to the moon, and other stuff like that. I wasn’t as directly involved in those—again, I was more of a low-TRL specialist—but some of my friends there were, so I became at least vaguely aware of the procedures and checks and balances that they were using.
I once wrote a blog post reviewing a crackpot physics book, in which I wrote the following:
Anyway, Newton’s laws are in fact proper physics, hence there’s a right and a wrong answer to how to apply Newton’s laws in any given situation. Sometimes people do it wrong, and sometimes they only notice that fact after running the experiment. But when that happens, then they should be able to find a locally-invalid step involved in the original prediction.
The claim “There’s a right answer to the question of how to apply Newton’s laws in any given situation” is a claim that Newton’s laws are meaningful, not a claim that they’re true in the real world. (Cf. “correctly” applying Newton’s laws to predict Mercury perihelion precession.) As such, you can verify this first claim without even running any experiments. E.g. get a bunch of good physicists, and show them all a bunch of local derivation steps in Newtonian physics problems, where some of those steps are locally-valid and others are locally-invalid, and see if the experts can figure out which is which. (Maybe you could make it easier on the experts by providing an argument for each side, a la “debate”.)
If Newton’s laws are meaningful (and they are), then it is at least possible in principle to compare their predictions to experiments. Then there’s still a separate question of whether the community of scientists and engineers are in fact capable of reliably doing that.
Luckily, every day around the world, there are thousands of new examples of honest-to-goodness predictions using Newton’s laws that turned out to be correct on the first try. That observation conveniently validates both the hypothesis that Newton’s laws are meaningful at all, and the hypothesis that scientists and engineers are collectively up to the task of applying Newton’s laws correctly, at least when they really set their minds to it.
At my old job we would sometimes be in a situation where we really needed our physics calculation to agree with experiments the first time. It was always pretty harrowing. Lots of people checking each other’s work and so on. Even so, the customers were always very skeptical of such claims—and probably justifiably so.
What was your old job?
I worked for five years at Draper, an R&D lab mostly in the USA military-industrial complex. I did a lot of calculations of the form “if we build this kind of gadget (usually a sensor), will it to actually meet the performance requirement?”. I was usually involved at a very early stage, when there was a wide space of possible design decisions and we hadn’t committed to anything yet, let alone started prototyping etc. As the project proceeds, you ramp up to more and more detailed models that make fewer and fewer assumptions, and you start supplementing that with prototype data and so on… but meanwhile the project costs are growing exponentially and it becomes almost impossible to make any more big design changes. So the calculations needed to be as faithful as possible, right from the earliest BOTEC / spreadsheet stage. No factor-of-2π errors allowed!! I think I was good at it … or at least, if I screwed anything up, nobody ever told me. :)
The firm also had projects that for things like designing & building actual gadgets that would then actually get launched to the moon, and other stuff like that. I wasn’t as directly involved in those—again, I was more of a low-TRL specialist—but some of my friends there were, so I became at least vaguely aware of the procedures and checks and balances that they were using.