Philosophical question about Quantum Gravity

[actually an egineer]

I imagined all possible quantum gravity theories in one set, of which only one is correct, and further divide these theories into testable and non-testable. Question: how do we know if the correct theory is currently testable? If it is testable, then we need to come up with new theories until eventually we discover a testable one and it turns out to be correct, so it's a problem in theoretical physics. If it is not testable, like string theory might be correct but we can't test it, then it's a problem in experimental physics.

My view on this issue is that theoretical physicists "gamble" to try and discover a testable theory that is correct at the same time. I don't have a problem with that, but is that essentially what they are doing? On the other hand, we can't develop experiments for these theories, because we need very high energies. So is it possible that string theory is correct but it will be proven in 10 000 years because we don't have the technology or economy to develop an experiment to test it?

 

[atyy]

It could be correct, and maybe even 10,000 years will not be enough to develop the technology to test it.

 

[majortom]

Or maybe there are multiple "correct" models.
If uncertainty sets a limit on the amount of information in the universe, then it's entirely plausible that we simply do not have enough information to arrive at a single model.
On the other hand, a non deterministic model can convey the idea of multiple correct models in a single framework. But, in doing so, these models lose precision.
Personally, I think this trade off between conceptual precision and accuracy is related to this effect of comparing deterministic and non deterministic models.

 

[actually an egineer]

Or maybe there are multiple "correct" models.
If uncertainty sets a limit on the amount of information in the universe, then it's entirely plausible that we simply do not have enough information to arrive at a single model.
On the other hand, a non deterministic model can convey the idea of multiple correct models in a single framework. But, in doing so, these models lose precision.
Personally, I think this trade off between conceptual precision and accuracy is related to this effect of comparing deterministic and non deterministic models.

The only way we find out if we can or cannot arrive at a single model is to try our best to arrive at one. If there are multiple correct models, can we find at least one?

 

[majortom]

The only way we find out if we can or cannot arrive at a single model is to try our best to arrive at one. If there are multiple correct models, can we find at least one?

Science has just an appalling history when it comes to proposing correct physical models. It seems that every time someone proposes a correct model, someone else finds a flaw in it.
I think it's worth considering this in terms of precision and accuracy, and considering the effect of determinism and non determinism that I described.
I do not think it is possible to create a physical model of arbitrary precision and accuracy, and I think this is the root cause of science's fallibility.
So, if your definition of "correct" is "a model of arbitrary precision and accuracy", then I would say that's likely impossible.
If your definition of "correct" is "a model of well defined and maximal precision and accuracy", then I think you're on the right path.
I'm more interested in models that capture uncertainty rather than reject it.

 

[actually an egineer]

Science has just an appalling history when it comes to proposing correct physical models. It seems that every time someone proposes a correct model, someone else finds a flaw in it.
I think it's worth considering this in terms of precision and accuracy, and considering the effect of determinism and non determinism that I described.
I do not think it is possible to create a physical model of arbitrary precision and accuracy, and I think this is the root cause of science's fallibility.
So, if your definition of "correct" is "a model of arbitrary precision and accuracy", then I would say that's likely impossible.
If your definition of "correct" is "a model of well defined and minimal precision and accuracy", then I think you're on the right path.
I'm more interested in models that capture uncertainty rather than reject it.

When I say "correct" I mean this. Newton had a correct theory of gravity because it explained a wide range of phenomena, but not all, so it was sufficiently accurate for his time. Then new phenomena arose which couldn't be explained with Newton's theory of gravity, we needed a more accurate theory of general relativity. Now we can't explain black holes or the Big Bang with the theory of general relativity so we need a theory of quantum gravity. A "correct" theory of quantum gravity will be the one that explains these phenomena and is experimentally confirmed. And if after theory of quantum gravity there are new phenomena which can't be explained by it, then we again need a more accurate theory.

 

[majortom]

When I say "correct" I mean this. Newton had a correct theory of gravity because it explained a wide range of phenomena, but not all, so it was sufficiently accurate for his time. Then new phenomena arose which couldn't be explained with Newton's theory of gravity, we needed a more accurate theory of general relativity. Now we can't explain black holes or the Big Bang with the theory of general relativity so we need a theory of quantum gravity. A "correct" theory of quantum gravity will be the one that explains these phenomena and is experimentally confirmed. And if after theory of quantum gravity there are new phenomena which can't be explained by it, then we again need a more accurate theory.

I don't quite agree with how you're using the word correct, but sure.
The problem I see is that uncertainty sets a boundary on how small a scale we can really test. This has never happened before. Newton didn't have a theory which would have a lower and upper scale limit, but that appears to be the universe we live in, experimentally.
If you're asking about what's testable and what isn't, well, Heisenberg did a pretty good job answering that question already.
Personally, I expect a "correct" (according to your definition) model will retain the uncertainty principle and revise schrodinger's wave equation.

 

[PeroK]

I do not think it is possible to create a physical model of arbitrary precision and accuracy, and I think this is the root cause of science's fallibility.

Science's fallability is its strength. It does not rely on infallible people or texts

Science has just an appalling history when it comes to proposing correct physical models.

Appalling is subjective. Moreover, science is only as good or bad as human capability allows it to be. If the science we have is bad, then where could the good science possibly have come from?

 

[majortom]

Science's fallability is its strength. It does not rely on infallible people or texts

Appalling is subjective. Moreover, science is only as good or bad as human capability allows it to be. If the science we have is bad, then where could the good science possibly have come from?

Correct is not subjective, at least not how I use the term. I use that term for logic and math proofs.
Science is not that. So, yes, science is appallingly incorrect when compared to math and logic.
Also, science isn't going to be limited by human capability for much longer, you need to think about the physical limits instead. It may even happen that science becomes mostly unintelligible to humans...

 

[robphy]

My point of view is that there are a variety of approaches,
which regard some aspects of known physics (especially in relativity and quantum theory)
as more fundamental than others.
For example, maybe causality is more fundamental than the Lorentz group.
Maybe discrete structures are more fundamental than manifolds and lie groups.
Maybe one needs a bigger lie group? Or a hierachy of particles?
In addition, we may have to generalize or develop new mathematics
to disentangle aspects that are blurred together by symmetries.

Armed with an array of mathematical tools and techniques,
guided by their point of view and (actual or possible) experimental results and theoretical results (possibly from other groups),
different groups (some more popular and better-funded than others)
try to push the ideas as far they can
[How far can one get with a certain set of ideas?],
looking to explain and experimentally test what is observed.
For example, can causality be the basis of physics,
and imply the Lorentz group in certain situations?

Certainly, experimental and theoretical results would help guide the research,
and maybe eliminate some of the options.

Of course, many other discoveries and inventions can be by-products
of this arguably exotic and apparently impractical field of research.
(Maxwell's Equations are about 160 years old. Schrodinger's Equation is about 100 years old.
Back then, did folks envision the computing power in our ubiquitous mobile phones?
Or the technology of nanoparticles and molecular machines?)

My $0.02.

 

[paradisePhysicist]

Correct is not subjective, at least not how I use the term. I use that term for logic and math proofs.
Science is not that. So, yes, science is appallingly incorrect when compared to math and logic.
Also, science isn't going to be limited by human capability for much longer, you need to think about the physical limits instead. It may even happen that science becomes mostly unintelligible to humans...

There could be Musk arcades, at museums of science. You got to the museum of science, plug in brain to the arcade, learn the theory of everything, then unplug and come out of the arcade feeling ecstatic about it but unable to explain what it is.